Revert "s3: listObjectParts return ErrNoSuchUpload if does not exist"

This reverts commit 6cf2e7d493.

.github/workflows/binaries_dev.yml

@@ -20,12 +20,12 @@ jobs:
           assets: |
             weed-*
 
-  build_dev_linux:
+  build_dev_linux_windows:
     needs: cleanup
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        goos: [linux]
+        goos: [linux, windows]
         goarch: [amd64]
 
     steps:
@@ -68,7 +68,7 @@ jobs:
           asset_name: "weed-${{ env.BUILD_TIME }}-${{ matrix.goos }}-${{ matrix.goarch }}"
 
   build_dev_darwin:
-    needs: build_dev_linux
+    needs: build_dev_linux_windows
     runs-on: ubuntu-latest
     strategy:
       matrix:

README.md

@@ -41,6 +41,7 @@ Your support will be really appreciated by me and other supporters!
 - [Download Binaries for different platforms](https://github.com/chrislusf/seaweedfs/releases/latest)
 - [SeaweedFS on Slack](https://join.slack.com/t/seaweedfs/shared_invite/enQtMzI4MTMwMjU2MzA3LTEyYzZmZWYzOGQ3MDJlZWMzYmI0OTE4OTJiZjJjODBmMzUxNmYwODg0YjY3MTNlMjBmZDQ1NzQ5NDJhZWI2ZmY)
 - [SeaweedFS on Twitter](https://twitter.com/SeaweedFS)
+- [SeaweedFS on Telegram](https://t.me/Seaweedfs) 
 - [SeaweedFS Mailing List](https://groups.google.com/d/forum/seaweedfs)
 - [Wiki Documentation](https://github.com/chrislusf/seaweedfs/wiki)
 - [SeaweedFS White Paper](https://github.com/chrislusf/seaweedfs/wiki/SeaweedFS_Architecture.pdf)

docker/Dockerfile.rocksdb_large

@@ -9,7 +9,7 @@ ENV ROCKSDB_VERSION v6.22.1
 RUN cd /tmp && \
     git clone https://github.com/facebook/rocksdb.git /tmp/rocksdb --depth 1 --single-branch --branch $ROCKSDB_VERSION && \
     cd rocksdb && \
-    make static_lib && \
+    PORTABLE=1 make static_lib && \
     make install-static
 
 ENV CGO_CFLAGS "-I/tmp/rocksdb/include"
@@ -29,7 +29,7 @@ FROM alpine AS final
 LABEL author="Chris Lu"
 COPY --from=builder /go/bin/weed /usr/bin/
 RUN mkdir -p /etc/seaweedfs
-COPY --from=builder /go/src/github.com/chrislusf/seaweedfs/docker/filer.toml /etc/seaweedfs/filer.toml
+COPY --from=builder /go/src/github.com/chrislusf/seaweedfs/docker/filer_rocksdb.toml /etc/seaweedfs/filer.toml
 COPY --from=builder /go/src/github.com/chrislusf/seaweedfs/docker/entrypoint.sh /entrypoint.sh
 RUN apk add fuse snappy gflags
 
@@ -50,7 +50,7 @@ EXPOSE 8333
 # webdav server http port
 EXPOSE 7333
 
-RUN mkdir -p /data/filerldb2
+RUN mkdir -p /data/filer_rocksdb
 
 VOLUME /data
 

docker/filer_rocksdb.toml

@@ -0,0 +1,3 @@
+[rocksdb]
+enabled = true
+dir = "/data/filer_rocksdb"

go.mod

@@ -8,7 +8,6 @@ require (
 	cloud.google.com/go/storage v1.16.1
 	github.com/Azure/azure-pipeline-go v0.2.3
 	github.com/Azure/azure-storage-blob-go v0.14.0
-	github.com/DataDog/zstd v1.3.6-0.20190409195224-796139022798 // indirect
 	github.com/OneOfOne/xxhash v1.2.2
 	github.com/Shopify/sarama v1.23.1
 	github.com/aws/aws-sdk-go v1.35.3
@@ -163,20 +162,22 @@ require (
 	modernc.org/token v1.0.0 // indirect
 )
 
+require (
+	github.com/fluent/fluent-logger-golang v1.8.0
+	github.com/hanwen/go-fuse/v2 v2.1.0
+)
+
 require (
 	cloud.google.com/go/kms v1.0.0 // indirect
-	github.com/apache/thrift v0.15.0 // indirect
+	github.com/DataDog/zstd v1.3.6-0.20190409195224-796139022798 // indirect
 	github.com/d4l3k/messagediff v1.2.1 // indirect
-	github.com/fluent/fluent-logger-golang v1.8.0 // indirect
 	github.com/gogo/protobuf v1.3.2 // indirect
-	github.com/gorilla/websocket v1.4.2 // indirect
 	github.com/jcmturner/aescts/v2 v2.0.0 // indirect
 	github.com/jcmturner/dnsutils/v2 v2.0.0 // indirect
 	github.com/jcmturner/goidentity/v6 v6.0.1 // indirect
 	github.com/jcmturner/rpc/v2 v2.0.2 // indirect
 	github.com/mattn/go-runewidth v0.0.7 // indirect
 	github.com/mattn/go-sqlite3 v2.0.1+incompatible // indirect
-	github.com/miekg/dns v1.1.44 // indirect
 	github.com/philhofer/fwd v1.1.1 // indirect
 	github.com/spf13/pflag v1.0.5 // indirect
 	github.com/tinylib/msgp v1.1.6 // indirect

go.sum

@@ -124,8 +124,6 @@ github.com/alecthomas/units v0.0.0-20190924025748-f65c72e2690d/go.mod h1:rBZYJk5
 github.com/antihax/optional v1.0.0/go.mod h1:uupD/76wgC+ih3iEmQUL+0Ugr19nfwCT1kdvxnR2qWY=
 github.com/apache/thrift v0.12.0/go.mod h1:cp2SuWMxlEZw2r+iP2GNCdIi4C1qmUzdZFSVb+bacwQ=
 github.com/apache/thrift v0.13.0/go.mod h1:cp2SuWMxlEZw2r+iP2GNCdIi4C1qmUzdZFSVb+bacwQ=
-github.com/apache/thrift v0.15.0 h1:aGvdaR0v1t9XLgjtBYwxcBvBOTMqClzwE26CHOgjW1Y=
-github.com/apache/thrift v0.15.0/go.mod h1:PHK3hniurgQaNMZYaCLEqXKsYK8upmhPbmdP2FXSqgU=
 github.com/armon/circbuf v0.0.0-20150827004946-bbbad097214e/go.mod h1:3U/XgcO3hCbHZ8TKRvWD2dDTCfh9M9ya+I9JpbB7O8o=
 github.com/armon/consul-api v0.0.0-20180202201655-eb2c6b5be1b6/go.mod h1:grANhF5doyWs3UAsr3K4I6qtAmlQcZDesFNEHPZAzj8=
 github.com/armon/go-metrics v0.0.0-20180917152333-f0300d1749da/go.mod h1:Q73ZrmVTwzkszR9V5SSuryQ31EELlFMUz1kKyl939pY=
@@ -281,30 +279,6 @@ github.com/go-stack/stack v1.8.0 h1:5SgMzNM5HxrEjV0ww2lTmX6E2Izsfxas4+YHWRs3Lsk=
 github.com/go-stack/stack v1.8.0/go.mod h1:v0f6uXyyMGvRgIKkXu+yp6POWl0qKG85gN/melR3HDY=
 github.com/go-zookeeper/zk v1.0.2 h1:4mx0EYENAdX/B/rbunjlt5+4RTA/a9SMHBRuSKdGxPM=
 github.com/go-zookeeper/zk v1.0.2/go.mod h1:nOB03cncLtlp4t+UAkGSV+9beXP/akpekBwL+UX1Qcw=
-github.com/gobuffalo/attrs v0.0.0-20190224210810-a9411de4debd/go.mod h1:4duuawTqi2wkkpB4ePgWMaai6/Kc6WEz83bhFwpHzj0=
-github.com/gobuffalo/depgen v0.0.0-20190329151759-d478694a28d3/go.mod h1:3STtPUQYuzV0gBVOY3vy6CfMm/ljR4pABfrTeHNLHUY=
-github.com/gobuffalo/depgen v0.1.0/go.mod h1:+ifsuy7fhi15RWncXQQKjWS9JPkdah5sZvtHc2RXGlg=
-github.com/gobuffalo/envy v1.6.15/go.mod h1:n7DRkBerg/aorDM8kbduw5dN3oXGswK5liaSCx4T5NI=
-github.com/gobuffalo/envy v1.7.0/go.mod h1:n7DRkBerg/aorDM8kbduw5dN3oXGswK5liaSCx4T5NI=
-github.com/gobuffalo/flect v0.1.0/go.mod h1:d2ehjJqGOH/Kjqcoz+F7jHTBbmDb38yXA598Hb50EGs=
-github.com/gobuffalo/flect v0.1.1/go.mod h1:8JCgGVbRjJhVgD6399mQr4fx5rRfGKVzFjbj6RE/9UI=
-github.com/gobuffalo/flect v0.1.3/go.mod h1:8JCgGVbRjJhVgD6399mQr4fx5rRfGKVzFjbj6RE/9UI=
-github.com/gobuffalo/genny v0.0.0-20190329151137-27723ad26ef9/go.mod h1:rWs4Z12d1Zbf19rlsn0nurr75KqhYp52EAGGxTbBhNk=
-github.com/gobuffalo/genny v0.0.0-20190403191548-3ca520ef0d9e/go.mod h1:80lIj3kVJWwOrXWWMRzzdhW3DsrdjILVil/SFKBzF28=
-github.com/gobuffalo/genny v0.1.0/go.mod h1:XidbUqzak3lHdS//TPu2OgiFB+51Ur5f7CSnXZ/JDvo=
-github.com/gobuffalo/genny v0.1.1/go.mod h1:5TExbEyY48pfunL4QSXxlDOmdsD44RRq4mVZ0Ex28Xk=
-github.com/gobuffalo/gitgen v0.0.0-20190315122116-cc086187d211/go.mod h1:vEHJk/E9DmhejeLeNt7UVvlSGv3ziL+djtTr3yyzcOw=
-github.com/gobuffalo/gogen v0.0.0-20190315121717-8f38393713f5/go.mod h1:V9QVDIxsgKNZs6L2IYiGR8datgMhB577vzTDqypH360=
-github.com/gobuffalo/gogen v0.1.0/go.mod h1:8NTelM5qd8RZ15VjQTFkAW6qOMx5wBbW4dSCS3BY8gg=
-github.com/gobuffalo/gogen v0.1.1/go.mod h1:y8iBtmHmGc4qa3urIyo1shvOD8JftTtfcKi+71xfDNE=
-github.com/gobuffalo/logger v0.0.0-20190315122211-86e12af44bc2/go.mod h1:QdxcLw541hSGtBnhUc4gaNIXRjiDppFGaDqzbrBd3v8=
-github.com/gobuffalo/mapi v1.0.1/go.mod h1:4VAGh89y6rVOvm5A8fKFxYG+wIW6LO1FMTG9hnKStFc=
-github.com/gobuffalo/mapi v1.0.2/go.mod h1:4VAGh89y6rVOvm5A8fKFxYG+wIW6LO1FMTG9hnKStFc=
-github.com/gobuffalo/packd v0.0.0-20190315124812-a385830c7fc0/go.mod h1:M2Juc+hhDXf/PnmBANFCqx4DM3wRbgDvnVWeG2RIxq4=
-github.com/gobuffalo/packd v0.1.0/go.mod h1:M2Juc+hhDXf/PnmBANFCqx4DM3wRbgDvnVWeG2RIxq4=
-github.com/gobuffalo/packr/v2 v2.0.9/go.mod h1:emmyGweYTm6Kdper+iywB6YK5YzuKchGtJQZ0Odn4pQ=
-github.com/gobuffalo/packr/v2 v2.2.0/go.mod h1:CaAwI0GPIAv+5wKLtv8Afwl+Cm78K/I/VCm/3ptBN+0=
-github.com/gobuffalo/syncx v0.0.0-20190224160051-33c29581e754/go.mod h1:HhnNqWY95UYwwW3uSASeV7vtgYkT2t16hJgV3AEPUpw=
 github.com/gocql/gocql v0.0.0-20210707082121-9a3953d1826d h1:k544nNVphXK4Yt0FTduvOvCfJabEY/DMkdNw0zpCwBE=
 github.com/gocql/gocql v0.0.0-20210707082121-9a3953d1826d/go.mod h1:3gM2c4D3AnkISwBxGnMMsS8Oy4y2lhbPRsH4xnJrHG8=
 github.com/godbus/dbus/v5 v5.0.4/go.mod h1:xhWf0FNVPg57R7Z0UbKHbJfkEywrmjJnf7w5xrFpKfA=
@@ -431,8 +405,6 @@ github.com/gorilla/sessions v1.2.0 h1:S7P+1Hm5V/AT9cjEcUD5uDaQSX0OE577aCXgoaKpYb
 github.com/gorilla/sessions v1.2.0/go.mod h1:dk2InVEVJ0sfLlnXv9EAgkf6ecYs/i80K/zI+bUmuGM=
 github.com/gorilla/websocket v0.0.0-20170926233335-4201258b820c/go.mod h1:E7qHFY5m1UJ88s3WnNqhKjPHQ0heANvMoAMk2YaljkQ=
 github.com/gorilla/websocket v1.4.0/go.mod h1:E7qHFY5m1UJ88s3WnNqhKjPHQ0heANvMoAMk2YaljkQ=
-github.com/gorilla/websocket v1.4.2 h1:+/TMaTYc4QFitKJxsQ7Yye35DkWvkdLcvGKqM+x0Ufc=
-github.com/gorilla/websocket v1.4.2/go.mod h1:YR8l580nyteQvAITg2hZ9XVh4b55+EU/adAjf1fMHhE=
 github.com/grpc-ecosystem/go-grpc-middleware v1.0.0/go.mod h1:FiyG127CGDf3tlThmgyCl78X/SZQqEOJBCDaAfeWzPs=
 github.com/grpc-ecosystem/go-grpc-middleware v1.0.1-0.20190118093823-f849b5445de4/go.mod h1:FiyG127CGDf3tlThmgyCl78X/SZQqEOJBCDaAfeWzPs=
 github.com/grpc-ecosystem/go-grpc-middleware v1.1.0 h1:THDBEeQ9xZ8JEaCLyLQqXMMdRqNr0QAUJTIkQAUtFjg=
@@ -443,6 +415,10 @@ github.com/grpc-ecosystem/grpc-gateway v1.9.5/go.mod h1:vNeuVxBJEsws4ogUvrchl83t
 github.com/grpc-ecosystem/grpc-gateway v1.16.0/go.mod h1:BDjrQk3hbvj6Nolgz8mAMFbcEtjT1g+wF4CSlocrBnw=
 github.com/hailocab/go-hostpool v0.0.0-20160125115350-e80d13ce29ed h1:5upAirOpQc1Q53c0bnx2ufif5kANL7bfZWcc6VJWJd8=
 github.com/hailocab/go-hostpool v0.0.0-20160125115350-e80d13ce29ed/go.mod h1:tMWxXQ9wFIaZeTI9F+hmhFiGpFmhOHzyShyFUhRm0H4=
+github.com/hanwen/go-fuse v1.0.0 h1:GxS9Zrn6c35/BnfiVsZVWmsG803xwE7eVRDvcf/BEVc=
+github.com/hanwen/go-fuse v1.0.0/go.mod h1:unqXarDXqzAk0rt98O2tVndEPIpUgLD9+rwFisZH3Ok=
+github.com/hanwen/go-fuse/v2 v2.1.0 h1:+32ffteETaLYClUj0a3aHjZ1hOPxxaNEHiZiujuDaek=
+github.com/hanwen/go-fuse/v2 v2.1.0/go.mod h1:oRyA5eK+pvJyv5otpO/DgccS8y/RvYMaO00GgRLGryc=
 github.com/hashicorp/consul/api v1.3.0/go.mod h1:MmDNSzIMUjNpY/mQ398R4bk2FnqQLoPndWW5VkKPlCE=
 github.com/hashicorp/consul/sdk v0.3.0/go.mod h1:VKf9jXwCTEY1QZP2MOLRhb5i/I/ssyNV1vwHyQBF0x8=
 github.com/hashicorp/errwrap v1.0.0 h1:hLrqtEDnRye3+sgx6z4qVLNuviH3MR5aQ0ykNJa/UYA=
@@ -516,16 +492,11 @@ github.com/karlseguin/ccache/v2 v2.0.7 h1:y5Pfi4eiyYCOD6LS/Kj+o6Nb4M5Ngpw9qFQs+v
 github.com/karlseguin/ccache/v2 v2.0.7/go.mod h1:2BDThcfQMf/c0jnZowt16eW405XIqZPavt+HoYEtcxQ=
 github.com/karlseguin/expect v1.0.2-0.20190806010014-778a5f0c6003 h1:vJ0Snvo+SLMY72r5J4sEfkuE7AFbixEP2qRbEcum/wA=
 github.com/karlseguin/expect v1.0.2-0.20190806010014-778a5f0c6003/go.mod h1:zNBxMY8P21owkeogJELCLeHIt+voOSduHYTFUbwRAV8=
-github.com/karrick/godirwalk v1.8.0/go.mod h1:H5KPZjojv4lE+QYImBI8xVtrBRgYrIVsaRPx4tDPEn4=
-github.com/karrick/godirwalk v1.10.3/go.mod h1:RoGL9dQei4vP9ilrpETWE8CLOZ1kiN0LhBygSwrAsHA=
 github.com/kballard/go-shellquote v0.0.0-20180428030007-95032a82bc51 h1:Z9n2FFNUXsshfwJMBgNA0RU6/i7WVaAegv3PtuIHPMs=
 github.com/kballard/go-shellquote v0.0.0-20180428030007-95032a82bc51/go.mod h1:CzGEWj7cYgsdH8dAjBGEr58BoE7ScuLd+fwFZ44+/x8=
 github.com/kisielk/errcheck v1.1.0/go.mod h1:EZBBE59ingxPouuu3KfxchcWSUPOHkagtvWXihfKN4Q=
 github.com/kisielk/errcheck v1.5.0/go.mod h1:pFxgyoBC7bSaBwPgfKdkLd5X25qrDl4LWUI2bnpBCr8=
 github.com/kisielk/gotool v1.0.0/go.mod h1:XhKaO+MFFWcvkIS/tQcRk01m1F5IRFswLeQ+oQHNcck=
-github.com/klauspost/compress v1.9.5/go.mod h1:RyIbtBH6LamlWaDj8nUwkbUhJ87Yi3uG0guNDohfE1A=
-github.com/klauspost/compress v1.10.9 h1:pPRt1Z78crspaHISkpSSHjDlx+Tt9suHe519dsI0vF4=
-github.com/klauspost/compress v1.10.9/go.mod h1:aoV0uJVorq1K+umq18yTdKaF57EivdYsUV+/s2qKfXs=
 github.com/klauspost/compress v1.13.6 h1:P76CopJELS0TiO2mebmnzgWaajssP/EszplttgQxcgc=
 github.com/klauspost/compress v1.13.6/go.mod h1:/3/Vjq9QcHkK5uEr5lBEmyoZ1iFhe47etQ6QUkpK6sk=
 github.com/klauspost/cpuid v1.2.1 h1:vJi+O/nMdFt0vqm8NZBI6wzALWdA2X+egi0ogNyrC/w=
@@ -533,7 +504,6 @@ github.com/klauspost/cpuid v1.2.1/go.mod h1:Pj4uuM528wm8OyEC2QMXAi2YiTZ96dNQPGgo
 github.com/klauspost/reedsolomon v1.9.2 h1:E9CMS2Pqbv+C7tsrYad4YC9MfhnMVWhMRsTi7U0UB18=
 github.com/klauspost/reedsolomon v1.9.2/go.mod h1:CwCi+NUr9pqSVktrkN+Ondf06rkhYZ/pcNv7fu+8Un4=
 github.com/konsorten/go-windows-terminal-sequences v1.0.1/go.mod h1:T0+1ngSBFLxvqU3pZ+m/2kptfBszLMUkC4ZK/EgS/cQ=
-github.com/konsorten/go-windows-terminal-sequences v1.0.2/go.mod h1:T0+1ngSBFLxvqU3pZ+m/2kptfBszLMUkC4ZK/EgS/cQ=
 github.com/konsorten/go-windows-terminal-sequences v1.0.3 h1:CE8S1cTafDpPvMhIxNJKvHsGVBgn1xWYf1NbHQhywc8=
 github.com/konsorten/go-windows-terminal-sequences v1.0.3/go.mod h1:T0+1ngSBFLxvqU3pZ+m/2kptfBszLMUkC4ZK/EgS/cQ=
 github.com/kr/fs v0.1.0/go.mod h1:FFnZGqtBN9Gxj7eW1uZ42v5BccTP0vu6NEaFoC2HwRg=
@@ -546,6 +516,8 @@ github.com/kr/text v0.1.0 h1:45sCR5RtlFHMR4UwH9sdQ5TC8v0qDQCHnXt+kaKSTVE=
 github.com/kr/text v0.1.0/go.mod h1:4Jbv+DJW3UT/LiOwJeYQe1efqtUx/iVham/4vfdArNI=
 github.com/kurin/blazer v0.5.3 h1:SAgYv0TKU0kN/ETfO5ExjNAPyMt2FocO2s/UlCHfjAk=
 github.com/kurin/blazer v0.5.3/go.mod h1:4FCXMUWo9DllR2Do4TtBd377ezyAJ51vB5uTBjt0pGU=
+github.com/kylelemons/godebug v0.0.0-20170820004349-d65d576e9348 h1:MtvEpTB6LX3vkb4ax0b5D2DHbNAUsen0Gx5wZoq3lV4=
+github.com/kylelemons/godebug v0.0.0-20170820004349-d65d576e9348/go.mod h1:B69LEHPfb2qLo0BaaOLcbitczOKLWTsrBG9LczfCD4k=
 github.com/lib/pq v1.1.1/go.mod h1:5WUZQaWbwv1U+lTReE5YruASi9Al49XbQIvNi/34Woo=
 github.com/lib/pq v1.10.0 h1:Zx5DJFEYQXio93kgXnQ09fXNiUKsqv4OUEu2UtGcB1E=
 github.com/lib/pq v1.10.0/go.mod h1:AlVN5x4E4T544tWzH6hKfbfQvm3HdbOxrmggDNAPY9o=
@@ -559,8 +531,6 @@ github.com/magiconair/properties v1.8.1 h1:ZC2Vc7/ZFkGmsVC9KvOjumD+G5lXy2RtTKyzR
 github.com/magiconair/properties v1.8.1/go.mod h1:PppfXfuXeibc/6YijjN8zIbojt8czPbwD3XqdrwzmxQ=
 github.com/mailru/easyjson v0.7.1 h1:mdxE1MF9o53iCb2Ghj1VfWvh7ZOwHpnVG/xwXrV90U8=
 github.com/mailru/easyjson v0.7.1/go.mod h1:KAzv3t3aY1NaHWoQz1+4F1ccyAH66Jk7yos7ldAVICs=
-github.com/markbates/oncer v0.0.0-20181203154359-bf2de49a0be2/go.mod h1:Ld9puTsIW75CHf65OeIOkyKbteujpZVXDpWK6YGZbxE=
-github.com/markbates/safe v1.0.1/go.mod h1:nAqgmRi7cY2nqMc92/bSEeQA+R4OheNU2T1kNSCBdG0=
 github.com/mattn/go-colorable v0.0.9/go.mod h1:9vuHe8Xs5qXnSaW/c/ABM9alt+Vo+STaOChaDxuIBZU=
 github.com/mattn/go-ieproxy v0.0.0-20190610004146-91bb50d98149/go.mod h1:31jz6HNzdxOmlERGGEc4v/dMssOfmp2p5bT/okiKFFc=
 github.com/mattn/go-ieproxy v0.0.0-20190702010315-6dee0af9227d/go.mod h1:31jz6HNzdxOmlERGGEc4v/dMssOfmp2p5bT/okiKFFc=
@@ -580,8 +550,6 @@ github.com/mattn/go-sqlite3 v2.0.1+incompatible/go.mod h1:FPy6KqzDD04eiIsT53CuJW
 github.com/matttproud/golang_protobuf_extensions v1.0.1 h1:4hp9jkHxhMHkqkrB3Ix0jegS5sx/RkqARlsWZ6pIwiU=
 github.com/matttproud/golang_protobuf_extensions v1.0.1/go.mod h1:D8He9yQNgCq6Z5Ld7szi9bcBfOoFv/3dc6xSMkL2PC0=
 github.com/miekg/dns v1.0.14/go.mod h1:W1PPwlIAgtquWBMBEV9nkV9Cazfe8ScdGz/Lj7v3Nrg=
-github.com/miekg/dns v1.1.44 h1:4rpqcegYPVkvIeVhITrKP1sRR3KjfRc1nrOPMUZmLyc=
-github.com/miekg/dns v1.1.44/go.mod h1:e3IlAVfNqAllflbibAZEWOXOQ+Ynzk/dDozDxY7XnME=
 github.com/mitchellh/cli v1.0.0/go.mod h1:hNIlj7HEI86fIcpObd7a0FcrxTWetlwJDGcceTlRvqc=
 github.com/mitchellh/go-homedir v1.0.0/go.mod h1:SfyaCUpYCn1Vlf4IUYiD9fPX4A5wJrkLzIz1N1q0pr0=
 github.com/mitchellh/go-homedir v1.1.0/go.mod h1:SfyaCUpYCn1Vlf4IUYiD9fPX4A5wJrkLzIz1N1q0pr0=
@@ -721,17 +689,11 @@ github.com/remyoudompheng/bigfft v0.0.0-20200410134404-eec4a21b6bb0 h1:OdAsTTz6O
 github.com/remyoudompheng/bigfft v0.0.0-20200410134404-eec4a21b6bb0/go.mod h1:qqbHyh8v60DhA7CoWK5oRCqLrMHRGoxYCSS9EjAz6Eo=
 github.com/rogpeppe/fastuuid v0.0.0-20150106093220-6724a57986af/go.mod h1:XWv6SoW27p1b0cqNHllgS5HIMJraePCO15w5zCzIWYg=
 github.com/rogpeppe/fastuuid v1.2.0/go.mod h1:jVj6XXZzXRy/MSR5jhDC/2q6DgLz+nrA6LYCDYWNEvQ=
-github.com/rogpeppe/go-internal v1.1.0/go.mod h1:M8bDsm7K2OlrFYOpmOWEs/qY81heoFRclV5y23lUDJ4=
-github.com/rogpeppe/go-internal v1.2.2/go.mod h1:M8bDsm7K2OlrFYOpmOWEs/qY81heoFRclV5y23lUDJ4=
 github.com/rogpeppe/go-internal v1.3.0/go.mod h1:M8bDsm7K2OlrFYOpmOWEs/qY81heoFRclV5y23lUDJ4=
 github.com/russross/blackfriday/v2 v2.0.1/go.mod h1:+Rmxgy9KzJVeS9/2gXHxylqXiyQDYRxCVz55jmeOWTM=
 github.com/ryanuber/columnize v0.0.0-20160712163229-9b3edd62028f/go.mod h1:sm1tb6uqfes/u+d4ooFouqFdy9/2g9QGwK3SQygK0Ts=
 github.com/samuel/go-zookeeper v0.0.0-20190923202752-2cc03de413da/go.mod h1:gi+0XIa01GRL2eRQVjQkKGqKF3SF9vZR/HnPullcV2E=
 github.com/sean-/seed v0.0.0-20170313163322-e2103e2c3529/go.mod h1:DxrIzT+xaE7yg65j358z/aeFdxmN0P9QXhEzd20vsDc=
-github.com/seaweedfs/fuse v1.2.0 h1:m4Ar3I0TY/Tzyb80lfJiXUKFcMBHqA5e/cZwxQtilZo=
-github.com/seaweedfs/fuse v1.2.0/go.mod h1:iwbDQv5BZACY54r6AO/6xsLNuMaYcBKSkLTZVfmK594=
-github.com/seaweedfs/fuse v1.2.1 h1:920YH3+PnIqpNLo2lsZDTUJHUPLXnxgQGEXoVb/yMno=
-github.com/seaweedfs/fuse v1.2.1/go.mod h1:iwbDQv5BZACY54r6AO/6xsLNuMaYcBKSkLTZVfmK594=
 github.com/seaweedfs/fuse v1.2.2 h1:01l8OjIdyATRNqVc/gDPgFobuC8ubQF3hRKOPColROw=
 github.com/seaweedfs/fuse v1.2.2/go.mod h1:iwbDQv5BZACY54r6AO/6xsLNuMaYcBKSkLTZVfmK594=
 github.com/seaweedfs/goexif v1.0.2 h1:p+rTXYdQ2mgxd+1JaTrQ9N8DvYuw9UH9xgYmJ+Bb29E=
@@ -740,8 +702,6 @@ github.com/secsy/goftp v0.0.0-20190720192957-f31499d7c79a h1:C6IhVTxNkhlb0tlCB6J
 github.com/secsy/goftp v0.0.0-20190720192957-f31499d7c79a/go.mod h1:MnkX001NG75g3p8bhFycnyIjeQoOjGL6CEIsdE/nKSY=
 github.com/shurcooL/sanitized_anchor_name v1.0.0/go.mod h1:1NzhyTcUVG4SuEtjjoZeVRXNmyL/1OwPU0+IJeTBvfc=
 github.com/sirupsen/logrus v1.2.0/go.mod h1:LxeOpSwHxABJmUn/MG1IvRgCAasNZTLOkJPxbbu5VWo=
-github.com/sirupsen/logrus v1.4.0/go.mod h1:LxeOpSwHxABJmUn/MG1IvRgCAasNZTLOkJPxbbu5VWo=
-github.com/sirupsen/logrus v1.4.1/go.mod h1:ni0Sbl8bgC9z8RoU9G6nDWqqs/fq4eDPysMBDgk/93Q=
 github.com/sirupsen/logrus v1.4.2/go.mod h1:tLMulIdttU9McNUspp0xgXVQah82FyeX6MwdIuYE2rE=
 github.com/sirupsen/logrus v1.5.0/go.mod h1:+F7Ogzej0PZc/94MaYx/nvG9jOFMD2osvC3s+Squfpo=
 github.com/sirupsen/logrus v1.6.0 h1:UBcNElsrwanuuMsnGSlYmtmgbb23qDR5dG+6X6Oo89I=
@@ -792,17 +752,11 @@ github.com/stvp/tempredis v0.0.0-20181119212430-b82af8480203 h1:QVqDTf3h2WHt08Yu
 github.com/stvp/tempredis v0.0.0-20181119212430-b82af8480203/go.mod h1:oqN97ltKNihBbwlX8dLpwxCl3+HnXKV/R0e+sRLd9C8=
 github.com/syndtr/goleveldb v1.0.1-0.20190318030020-c3a204f8e965 h1:1oFLiOyVl+W7bnBzGhf7BbIv9loSFQcieWWYIjLqcAw=
 github.com/syndtr/goleveldb v1.0.1-0.20190318030020-c3a204f8e965/go.mod h1:9OrXJhf154huy1nPWmuSrkgjPUtUNhA+Zmy+6AESzuA=
-github.com/tidwall/gjson v1.8.1 h1:8j5EE9Hrh3l9Od1OIEDAb7IpezNA20UdRngNAj5N0WU=
-github.com/tidwall/gjson v1.8.1/go.mod h1:5/xDoumyyDNerp2U36lyolv46b3uF/9Bu6OfyQ9GImk=
 github.com/tidwall/gjson v1.10.2 h1:APbLGOM0rrEkd8WBw9C24nllro4ajFuJu0Sc9hRz8Bo=
 github.com/tidwall/gjson v1.10.2/go.mod h1:/wbyibRr2FHMks5tjHJ5F8dMZh3AcwJEMf5vlfC0lxk=
-github.com/tidwall/match v1.0.3 h1:FQUVvBImDutD8wJLN6c5eMzWtjgONK9MwIBCOrUJKeE=
-github.com/tidwall/match v1.0.3/go.mod h1:eRSPERbgtNPcGhD8UCthc6PmLEQXEWd3PRB5JTxsfmM=
 github.com/tidwall/match v1.1.1 h1:+Ho715JplO36QYgwN9PGYNhgZvoUSc9X2c80KVTi+GA=
 github.com/tidwall/match v1.1.1/go.mod h1:eRSPERbgtNPcGhD8UCthc6PmLEQXEWd3PRB5JTxsfmM=
 github.com/tidwall/pretty v1.0.0/go.mod h1:XNkn88O1ChpSDQmQeStsy+sBenx6DDtFZJxhVysOjyk=
-github.com/tidwall/pretty v1.1.0 h1:K3hMW5epkdAVwibsQEfR/7Zj0Qgt4DxtNumTq/VloO8=
-github.com/tidwall/pretty v1.1.0/go.mod h1:XNkn88O1ChpSDQmQeStsy+sBenx6DDtFZJxhVysOjyk=
 github.com/tidwall/pretty v1.2.0 h1:RWIZEg2iJ8/g6fDDYzMpobmaoGh5OLl4AXtGUGPcqCs=
 github.com/tidwall/pretty v1.2.0/go.mod h1:ITEVvHYasfjBbM0u2Pg8T2nJnzm8xPwvNhhsoaGGjNU=
 github.com/tinylib/msgp v1.1.6 h1:i+SbKraHhnrf9M5MYmvQhFnbLhAXSDWF8WWsuyRdocw=
@@ -853,8 +807,6 @@ go.etcd.io/etcd/client/pkg/v3 v3.5.0 h1:2aQv6F436YnN7I4VbI8PPYrBhu+SmrTaADcf8Mi/
 go.etcd.io/etcd/client/pkg/v3 v3.5.0/go.mod h1:IJHfcCEKxYu1Os13ZdwCwIUTUVGYTSAM3YSwc9/Ac1g=
 go.etcd.io/etcd/client/v3 v3.5.0 h1:62Eh0XOro+rDwkrypAGDfgmNh5Joq+z+W9HZdlXMzek=
 go.etcd.io/etcd/client/v3 v3.5.0/go.mod h1:AIKXXVX/DQXtfTEqBryiLTUXwON+GuvO6Z7lLS/oTh0=
-go.mongodb.org/mongo-driver v1.7.0 h1:hHrvOBWlWB2c7+8Gh/Xi5jj82AgidK/t7KVXBZ+IyUA=
-go.mongodb.org/mongo-driver v1.7.0/go.mod h1:Q4oFMbo1+MSNqICAdYMlC/zSTrwCogR4R8NzkI+yfU8=
 go.mongodb.org/mongo-driver v1.8.0 h1:R/P/JJzu8LJvJ1lDfph9GLNIKQxEtIHFfnUUUve35zY=
 go.mongodb.org/mongo-driver v1.8.0/go.mod h1:0sQWfOeY63QTntERDJJ/0SuKK0T1uVSgKCuAROlKEPY=
 go.opencensus.io v0.15.0/go.mod h1:UffZAU+4sDEINUGP/B7UfBBkq4fqLu9zXAX7ke6CHW0=
@@ -898,7 +850,6 @@ golang.org/x/crypto v0.0.0-20181029021203-45a5f77698d3/go.mod h1:6SG95UA2DQfeDnf
 golang.org/x/crypto v0.0.0-20181203042331-505ab145d0a9/go.mod h1:6SG95UA2DQfeDnfUPMdvaQW0Q7yPrPDi9nlGo2tz2b4=
 golang.org/x/crypto v0.0.0-20190308221718-c2843e01d9a2/go.mod h1:djNgcEr1/C05ACkg1iLfiJU5Ep61QUkGW8qpdssI0+w=
 golang.org/x/crypto v0.0.0-20190404164418-38d8ce5564a5/go.mod h1:WFFai1msRO1wXaEeE5yQxYXgSfI8pQAWXbQop6sCtWE=
-golang.org/x/crypto v0.0.0-20190422162423-af44ce270edf/go.mod h1:WFFai1msRO1wXaEeE5yQxYXgSfI8pQAWXbQop6sCtWE=
 golang.org/x/crypto v0.0.0-20190510104115-cbcb75029529/go.mod h1:yigFU9vqHzYiE8UmvKecakEJjdnWj3jj499lnFckfCI=
 golang.org/x/crypto v0.0.0-20190530122614-20be4c3c3ed5/go.mod h1:yigFU9vqHzYiE8UmvKecakEJjdnWj3jj499lnFckfCI=
 golang.org/x/crypto v0.0.0-20190605123033-f99c8df09eb5/go.mod h1:yigFU9vqHzYiE8UmvKecakEJjdnWj3jj499lnFckfCI=
@@ -907,7 +858,6 @@ golang.org/x/crypto v0.0.0-20190820162420-60c769a6c586/go.mod h1:yigFU9vqHzYiE8U
 golang.org/x/crypto v0.0.0-20191011191535-87dc89f01550/go.mod h1:yigFU9vqHzYiE8UmvKecakEJjdnWj3jj499lnFckfCI=
 golang.org/x/crypto v0.0.0-20191206172530-e9b2fee46413/go.mod h1:LzIPMQfyMNhhGPhUkYOs5KpL4U8rLKemX1yGLhDgUto=
 golang.org/x/crypto v0.0.0-20200117160349-530e935923ad/go.mod h1:LzIPMQfyMNhhGPhUkYOs5KpL4U8rLKemX1yGLhDgUto=
-golang.org/x/crypto v0.0.0-20200302210943-78000ba7a073/go.mod h1:LzIPMQfyMNhhGPhUkYOs5KpL4U8rLKemX1yGLhDgUto=
 golang.org/x/crypto v0.0.0-20200323165209-0ec3e9974c59/go.mod h1:LzIPMQfyMNhhGPhUkYOs5KpL4U8rLKemX1yGLhDgUto=
 golang.org/x/crypto v0.0.0-20200604202706-70a84ac30bf9/go.mod h1:LzIPMQfyMNhhGPhUkYOs5KpL4U8rLKemX1yGLhDgUto=
 golang.org/x/crypto v0.0.0-20200622213623-75b288015ac9/go.mod h1:LzIPMQfyMNhhGPhUkYOs5KpL4U8rLKemX1yGLhDgUto=
@@ -1003,7 +953,6 @@ golang.org/x/net v0.0.0-20210226172049-e18ecbb05110/go.mod h1:m0MpNAwzfU5UDzcl9v
 golang.org/x/net v0.0.0-20210316092652-d523dce5a7f4/go.mod h1:RBQZq4jEuRlivfhVLdyRGr576XBO4/greRjx4P4O3yc=
 golang.org/x/net v0.0.0-20210405180319-a5a99cb37ef4/go.mod h1:p54w0d4576C0XHj96bSt6lcn1PtDYWL6XObtHCRCNQM=
 golang.org/x/net v0.0.0-20210503060351-7fd8e65b6420/go.mod h1:9nx3DQGgdP8bBQD5qxJ1jj9UTztislL4KSBs9R2vV5Y=
-golang.org/x/net v0.0.0-20210726213435-c6fcb2dbf985/go.mod h1:9nx3DQGgdP8bBQD5qxJ1jj9UTztislL4KSBs9R2vV5Y=
 golang.org/x/net v0.0.0-20210813160813-60bc85c4be6d h1:LO7XpTYMwTqxjLcGWPijK3vRXg1aWdlNOVOHRq45d7c=
 golang.org/x/net v0.0.0-20210813160813-60bc85c4be6d/go.mod h1:9nx3DQGgdP8bBQD5qxJ1jj9UTztislL4KSBs9R2vV5Y=
 golang.org/x/oauth2 v0.0.0-20180821212333-d2e6202438be/go.mod h1:N/0e6XlmueqKjAGxoOufVs8QHGRruUQn6yWY3a++T0U=
@@ -1027,7 +976,6 @@ golang.org/x/sync v0.0.0-20180314180146-1d60e4601c6f/go.mod h1:RxMgew5VJxzue5/jJ
 golang.org/x/sync v0.0.0-20181108010431-42b317875d0f/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
 golang.org/x/sync v0.0.0-20181221193216-37e7f081c4d4/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
 golang.org/x/sync v0.0.0-20190227155943-e225da77a7e6/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
-golang.org/x/sync v0.0.0-20190412183630-56d357773e84/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
 golang.org/x/sync v0.0.0-20190423024810-112230192c58/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
 golang.org/x/sync v0.0.0-20190911185100-cd5d95a43a6e/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
 golang.org/x/sync v0.0.0-20200317015054-43a5402ce75a/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
@@ -1048,11 +996,9 @@ golang.org/x/sys v0.0.0-20190215142949-d0b11bdaac8a/go.mod h1:STP8DvDyc/dI5b8T5h
 golang.org/x/sys v0.0.0-20190312061237-fead79001313/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20190403152447-81d4e9dc473e/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20190412213103-97732733099d/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
-golang.org/x/sys v0.0.0-20190419153524-e8e3143a4f4a/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20190422165155-953cdadca894/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20190502145724-3ef323f4f1fd/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20190507160741-ecd444e8653b/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
-golang.org/x/sys v0.0.0-20190531175056-4c3a928424d2/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20190606165138-5da285871e9c/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20190624142023-c5567b49c5d0/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20190726091711-fc99dfbffb4e/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
@@ -1112,10 +1058,7 @@ golang.org/x/sys v0.0.0-20210630005230-0f9fa26af87c/go.mod h1:oPkhp1MJrh7nUepCBc
 golang.org/x/sys v0.0.0-20210806184541-e5e7981a1069/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.0.0-20210817142637-7d9622a276b7/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.0.0-20210823070655-63515b42dcdf/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
-golang.org/x/sys v0.0.0-20210908233432-aa78b53d3365 h1:6wSTsvPddg9gc/mVEEyk9oOAoxn+bT4Z9q1zx+4RwA4=
 golang.org/x/sys v0.0.0-20210908233432-aa78b53d3365/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
-golang.org/x/sys v0.0.0-20211216021012-1d35b9e2eb4e h1:fLOSk5Q00efkSvAm+4xcoXD+RRmLmmulPn5I3Y9F2EM=
-golang.org/x/sys v0.0.0-20211216021012-1d35b9e2eb4e/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/sys v0.0.0-20220111092808-5a964db01320 h1:0jf+tOCoZ3LyutmCOWpVni1chK4VfFLhRsDK7MhqGRY=
 golang.org/x/sys v0.0.0-20220111092808-5a964db01320/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=
 golang.org/x/term v0.0.0-20201117132131-f5c789dd3221/go.mod h1:Nr5EML6q2oocZ2LXRh80K7BxOlk5/8JxuGnuhpl+muw=
@@ -1127,7 +1070,6 @@ golang.org/x/text v0.3.2/go.mod h1:bEr9sfX3Q8Zfm5fL9x+3itogRgK3+ptLWKqgva+5dAk=
 golang.org/x/text v0.3.3/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ=
 golang.org/x/text v0.3.4/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ=
 golang.org/x/text v0.3.5/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ=
-golang.org/x/text v0.3.6 h1:aRYxNxv6iGQlyVaZmk6ZgYEDa+Jg18DxebPSrd6bg1M=
 golang.org/x/text v0.3.6/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ=
 golang.org/x/text v0.3.7 h1:olpwvP2KacW1ZWvsR7uQhoyTYvKAupfQrRGBFM352Gk=
 golang.org/x/text v0.3.7/go.mod h1:u+2+/6zg+i71rQMx5EYifcz6MCKuco9NR6JIITiCfzQ=
@@ -1144,9 +1086,6 @@ golang.org/x/tools v0.0.0-20190311212946-11955173bddd/go.mod h1:LCzVGOaR6xXOjkQ3
 golang.org/x/tools v0.0.0-20190312151545-0bb0c0a6e846/go.mod h1:LCzVGOaR6xXOjkQ3onu1FJEFr0SW1gC7cKk1uF8kGRs=
 golang.org/x/tools v0.0.0-20190312170243-e65039ee4138/go.mod h1:LCzVGOaR6xXOjkQ3onu1FJEFr0SW1gC7cKk1uF8kGRs=
 golang.org/x/tools v0.0.0-20190328211700-ab21143f2384/go.mod h1:LCzVGOaR6xXOjkQ3onu1FJEFr0SW1gC7cKk1uF8kGRs=
-golang.org/x/tools v0.0.0-20190329151228-23e29df326fe/go.mod h1:LCzVGOaR6xXOjkQ3onu1FJEFr0SW1gC7cKk1uF8kGRs=
-golang.org/x/tools v0.0.0-20190416151739-9c9e1878f421/go.mod h1:LCzVGOaR6xXOjkQ3onu1FJEFr0SW1gC7cKk1uF8kGRs=
-golang.org/x/tools v0.0.0-20190420181800-aa740d480789/go.mod h1:LCzVGOaR6xXOjkQ3onu1FJEFr0SW1gC7cKk1uF8kGRs=
 golang.org/x/tools v0.0.0-20190422233926-fe54fb35175b/go.mod h1:LCzVGOaR6xXOjkQ3onu1FJEFr0SW1gC7cKk1uF8kGRs=
 golang.org/x/tools v0.0.0-20190425150028-36563e24a262/go.mod h1:RgjU9mgBXZiqYHBnxXauZ1Gv1EHHAz9KjViQ78xBX0Q=
 golang.org/x/tools v0.0.0-20190506145303-2d16b83fe98c/go.mod h1:RgjU9mgBXZiqYHBnxXauZ1Gv1EHHAz9KjViQ78xBX0Q=
@@ -1205,7 +1144,6 @@ golang.org/x/tools v0.1.1/go.mod h1:o0xws9oXOQQZyjljx8fwUC0k7L1pTE6eaCbjGeHmOkk=
 golang.org/x/tools v0.1.2/go.mod h1:o0xws9oXOQQZyjljx8fwUC0k7L1pTE6eaCbjGeHmOkk=
 golang.org/x/tools v0.1.3/go.mod h1:o0xws9oXOQQZyjljx8fwUC0k7L1pTE6eaCbjGeHmOkk=
 golang.org/x/tools v0.1.4/go.mod h1:o0xws9oXOQQZyjljx8fwUC0k7L1pTE6eaCbjGeHmOkk=
-golang.org/x/tools v0.1.5 h1:ouewzE6p+/VEB31YYnTbEJdi8pFqKp4P4n85vwo3DHA=
 golang.org/x/tools v0.1.5/go.mod h1:o0xws9oXOQQZyjljx8fwUC0k7L1pTE6eaCbjGeHmOkk=
 golang.org/x/tools v0.1.6-0.20210726203631-07bc1bf47fb2 h1:BonxutuHCTL0rBDnZlKjpGIQFTjyUVTexFOdWkB6Fg0=
 golang.org/x/tools v0.1.6-0.20210726203631-07bc1bf47fb2/go.mod h1:o0xws9oXOQQZyjljx8fwUC0k7L1pTE6eaCbjGeHmOkk=

k8s/helm_charts2/Chart.yaml

@@ -1,5 +1,5 @@
 apiVersion: v1
 description: SeaweedFS
 name: seaweedfs
-appVersion: "2.86"
-version: "2.86"
+appVersion: "2.88"
+version: "2.88"

k8s/helm_charts2/templates/volume-statefulset.yaml

@@ -40,7 +40,7 @@ spec:
       imagePullSecrets:
         - name: {{ .Values.global.imagePullSecrets }}
       {{- end }}
-      terminationGracePeriodSeconds: 10
+      terminationGracePeriodSeconds: 150
       {{- if .Values.volume.priorityClassName }}
       priorityClassName: {{ .Values.volume.priorityClassName | quote }}
       {{- end }}

other/java/client/src/main/proto/filer.proto

@@ -114,6 +114,7 @@ message Entry {
     bytes content = 9; // if not empty, the file content
 
     RemoteEntry remote_entry = 10;
+    int64 quota = 11; // for bucket only. Positive/Negative means enabled/disabled.
 }
 
 message FullEntry {

other/java/examples/pom.xml

@@ -23,7 +23,7 @@
         <dependency>
             <groupId>org.apache.hadoop</groupId>
             <artifactId>hadoop-common</artifactId>
-            <version>2.9.2</version>
+            <version>2.10.1</version>
             <scope>compile</scope>
         </dependency>
     </dependencies>

other/java/hdfs-over-ftp/pom.xml

@@ -36,7 +36,7 @@
         <dependency>
             <groupId>org.apache.hadoop</groupId>
             <artifactId>hadoop-common</artifactId>
-            <version>3.2.1</version>
+            <version>3.2.2</version>
         </dependency>
         <dependency>
             <groupId>org.apache.hadoop</groupId>

other/java/hdfs2/pom.xml

@@ -6,7 +6,7 @@
 
     <properties>
         <seaweedfs.client.version>2.85</seaweedfs.client.version>
-        <hadoop.version>2.9.2</hadoop.version>
+        <hadoop.version>2.10.1</hadoop.version>
     </properties>
 
     <groupId>com.github.chrislusf</groupId>

other/java/hdfs3/pom.xml

@@ -6,7 +6,7 @@
 
     <properties>
         <seaweedfs.client.version>2.85</seaweedfs.client.version>
-        <hadoop.version>3.1.1</hadoop.version>
+        <hadoop.version>3.1.4</hadoop.version>
     </properties>
 
     <groupId>com.github.chrislusf</groupId>

weed/Makefile

@@ -21,6 +21,10 @@ debug_mount:
 	go build -gcflags="all=-N -l"
 	dlv --listen=:2345 --headless=true --api-version=2 --accept-multiclient exec weed -- -v=4 mount -dir=~/tmp/mm -cacheCapacityMB=0 -filer.path=/ -umask=000
 
+debug_mount2:
+	go build -gcflags="all=-N -l"
+	dlv --listen=:2345 --headless=true --api-version=2 --accept-multiclient exec weed -- -v=4 mount2 -dir=~/tmp/mm -cacheCapacityMB=0 -filer.path=/ -umask=000
+
 debug_server:
 	go build -gcflags="all=-N -l"
 	dlv --listen=:2345 --headless=true --api-version=2 --accept-multiclient exec weed -- server -dir=~/tmp/99 -filer -volume.port=8343 -s3 -volume.max=0 -master.volumeSizeLimitMB=1024 -volume.preStopSeconds=1

weed/command/command.go

@@ -30,6 +30,7 @@ var Commands = []*Command{
 	cmdMaster,
 	cmdMasterFollower,
 	cmdMount,
+	cmdMount2,
 	cmdS3,
 	cmdIam,
 	cmdMsgBroker,

weed/command/filer_cat.go

@@ -8,7 +8,6 @@ import (
 	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
 	"github.com/chrislusf/seaweedfs/weed/wdclient"
 	"google.golang.org/grpc"
-	"math"
 	"net/url"
 	"os"
 	"strings"
@@ -115,7 +114,7 @@ func runFilerCat(cmd *Command, args []string) bool {
 
 		filerCat.filerClient = client
 
-		return filer.StreamContent(&filerCat, writer, respLookupEntry.Entry.Chunks, 0, math.MaxInt64)
+		return filer.StreamContent(&filerCat, writer, respLookupEntry.Entry.Chunks, 0, int64(filer.FileSize(respLookupEntry.Entry)))
 
 	})
 

weed/command/filer_copy.go

@@ -212,7 +212,7 @@ func genFileCopyTask(fileOrDir string, destPath string, fileCopyTaskChan chan Fi
 	if mode.IsDir() {
 		files, _ := os.ReadDir(fileOrDir)
 		for _, subFileOrDir := range files {
-			cleanedDestDirectory := filepath.Clean(destPath + fi.Name())
+			cleanedDestDirectory := destPath + fi.Name()
 			if err = genFileCopyTask(fileOrDir+"/"+subFileOrDir.Name(), cleanedDestDirectory+"/", fileCopyTaskChan); err != nil {
 				return err
 			}

weed/command/filer_remote_gateway_buckets.go

@@ -264,7 +264,7 @@ func (option *RemoteGatewayOptions) makeBucketedEventProcessor(filerSource *sour
 						// update directory property
 						return nil
 					}
-					if filer.IsSameData(message.OldEntry, message.NewEntry) {
+					if message.OldEntry.RemoteEntry != nil && filer.IsSameData(message.OldEntry, message.NewEntry) {
 						glog.V(2).Infof("update meta: %+v", resp)
 						oldDest := toRemoteStorageLocation(oldBucket, util.NewFullPath(resp.Directory, message.OldEntry.Name), oldRemoteStorageMountLocation)
 						return client.UpdateFileMetadata(oldDest, message.OldEntry, message.NewEntry)

weed/command/master.go

@@ -2,6 +2,7 @@ package command
 
 import (
 	"github.com/chrislusf/raft/protobuf"
+	stats_collect "github.com/chrislusf/seaweedfs/weed/stats"
 	"github.com/gorilla/mux"
 	"google.golang.org/grpc/reflection"
 	"net/http"
@@ -42,6 +43,7 @@ type MasterOptions struct {
 	metricsAddress     *string
 	metricsIntervalSec *int
 	raftResumeState    *bool
+	metricsHttpPort    *int
 }
 
 func init() {
@@ -61,6 +63,7 @@ func init() {
 	m.disableHttp = cmdMaster.Flag.Bool("disableHttp", false, "disable http requests, only gRPC operations are allowed.")
 	m.metricsAddress = cmdMaster.Flag.String("metrics.address", "", "Prometheus gateway address <host>:<port>")
 	m.metricsIntervalSec = cmdMaster.Flag.Int("metrics.intervalSeconds", 15, "Prometheus push interval in seconds")
+	m.metricsHttpPort = cmdMaster.Flag.Int("metricsPort", 0, "Prometheus metrics listen port")
 	m.raftResumeState = cmdMaster.Flag.Bool("resumeState", false, "resume previous state on start master server")
 }
 
@@ -104,6 +107,7 @@ func runMaster(cmd *Command, args []string) bool {
 		glog.Fatalf("volumeSizeLimitMB should be smaller than 30000")
 	}
 
+	go stats_collect.StartMetricsServer(*m.metricsHttpPort)
 	startMaster(m, masterWhiteList)
 
 	return true

weed/command/mount2.go

@@ -0,0 +1,83 @@
+package command
+
+import (
+	"os"
+	"time"
+)
+
+type Mount2Options struct {
+	filer              *string
+	filerMountRootPath *string
+	dir                *string
+	dirAutoCreate      *bool
+	collection         *string
+	replication        *string
+	diskType           *string
+	ttlSec             *int
+	chunkSizeLimitMB   *int
+	concurrentWriters  *int
+	cacheDir           *string
+	cacheSizeMB        *int64
+	dataCenter         *string
+	allowOthers        *bool
+	umaskString        *string
+	nonempty           *bool
+	volumeServerAccess *string
+	uidMap             *string
+	gidMap             *string
+	readOnly           *bool
+	debug              *bool
+	debugPort          *int
+}
+
+var (
+	mount2Options Mount2Options
+)
+
+func init() {
+	cmdMount2.Run = runMount2 // break init cycle
+	mount2Options.filer = cmdMount2.Flag.String("filer", "localhost:8888", "comma-separated weed filer location")
+	mount2Options.filerMountRootPath = cmdMount2.Flag.String("filer.path", "/", "mount this remote path from filer server")
+	mount2Options.dir = cmdMount2.Flag.String("dir", ".", "mount weed filer to this directory")
+	mount2Options.dirAutoCreate = cmdMount2.Flag.Bool("dirAutoCreate", false, "auto create the directory to mount to")
+	mount2Options.collection = cmdMount2.Flag.String("collection", "", "collection to create the files")
+	mount2Options.replication = cmdMount2.Flag.String("replication", "", "replication(e.g. 000, 001) to create to files. If empty, let filer decide.")
+	mount2Options.diskType = cmdMount2.Flag.String("disk", "", "[hdd|ssd|<tag>] hard drive or solid state drive or any tag")
+	mount2Options.ttlSec = cmdMount2.Flag.Int("ttl", 0, "file ttl in seconds")
+	mount2Options.chunkSizeLimitMB = cmdMount2.Flag.Int("chunkSizeLimitMB", 2, "local write buffer size, also chunk large files")
+	mount2Options.concurrentWriters = cmdMount2.Flag.Int("concurrentWriters", 32, "limit concurrent goroutine writers if not 0")
+	mount2Options.cacheDir = cmdMount2.Flag.String("cacheDir", os.TempDir(), "local cache directory for file chunks and meta data")
+	mount2Options.cacheSizeMB = cmdMount2.Flag.Int64("cacheCapacityMB", 1000, "local file chunk cache capacity in MB (0 will disable cache)")
+	mount2Options.dataCenter = cmdMount2.Flag.String("dataCenter", "", "prefer to write to the data center")
+	mount2Options.allowOthers = cmdMount2.Flag.Bool("allowOthers", true, "allows other users to access the file system")
+	mount2Options.umaskString = cmdMount2.Flag.String("umask", "022", "octal umask, e.g., 022, 0111")
+	mount2Options.nonempty = cmdMount2.Flag.Bool("nonempty", false, "allows the mounting over a non-empty directory")
+	mount2Options.volumeServerAccess = cmdMount2.Flag.String("volumeServerAccess", "direct", "access volume servers by [direct|publicUrl|filerProxy]")
+	mount2Options.uidMap = cmdMount2.Flag.String("map.uid", "", "map local uid to uid on filer, comma-separated <local_uid>:<filer_uid>")
+	mount2Options.gidMap = cmdMount2.Flag.String("map.gid", "", "map local gid to gid on filer, comma-separated <local_gid>:<filer_gid>")
+	mount2Options.readOnly = cmdMount2.Flag.Bool("readOnly", false, "read only")
+	mount2Options.debug = cmdMount2.Flag.Bool("debug", false, "serves runtime profiling data, e.g., http://localhost:<debug.port>/debug/pprof/goroutine?debug=2")
+	mount2Options.debugPort = cmdMount2.Flag.Int("debug.port", 6061, "http port for debugging")
+
+	mountCpuProfile = cmdMount2.Flag.String("cpuprofile", "", "cpu profile output file")
+	mountMemProfile = cmdMount2.Flag.String("memprofile", "", "memory profile output file")
+	mountReadRetryTime = cmdMount2.Flag.Duration("readRetryTime", 6*time.Second, "maximum read retry wait time")
+}
+
+var cmdMount2 = &Command{
+	UsageLine: "mount2 -filer=localhost:8888 -dir=/some/dir",
+	Short:     "<WIP> mount weed filer to a directory as file system in userspace(FUSE)",
+	Long: `mount weed filer to userspace.
+
+  Pre-requisites:
+  1) have SeaweedFS master and volume servers running
+  2) have a "weed filer" running
+  These 2 requirements can be achieved with one command "weed server -filer=true"
+
+  This uses github.com/seaweedfs/fuse, which enables writing FUSE file systems on
+  Linux, and OS X.
+
+  On OS X, it requires OSXFUSE (http://osxfuse.github.com/).
+
+  `,
+}

weed/command/mount2_std.go

@@ -0,0 +1,210 @@
+package command
+
+import (
+	"context"
+	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/mount"
+	"github.com/chrislusf/seaweedfs/weed/mount/meta_cache"
+	"github.com/chrislusf/seaweedfs/weed/mount/unmount"
+	"github.com/chrislusf/seaweedfs/weed/pb"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/chrislusf/seaweedfs/weed/security"
+	"github.com/chrislusf/seaweedfs/weed/storage/types"
+	"github.com/hanwen/go-fuse/v2/fuse"
+	"net/http"
+	"os"
+	"os/user"
+	"runtime"
+	"strconv"
+	"strings"
+	"time"
+
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"github.com/chrislusf/seaweedfs/weed/util/grace"
+)
+
+func runMount2(cmd *Command, args []string) bool {
+
+	if *mountOptions.debug {
+		go http.ListenAndServe(fmt.Sprintf(":%d", *mountOptions.debugPort), nil)
+	}
+
+	grace.SetupProfiling(*mountCpuProfile, *mountMemProfile)
+	if *mountReadRetryTime < time.Second {
+		*mountReadRetryTime = time.Second
+	}
+	util.RetryWaitTime = *mountReadRetryTime
+
+	umask, umaskErr := strconv.ParseUint(*mountOptions.umaskString, 8, 64)
+	if umaskErr != nil {
+		fmt.Printf("can not parse umask %s", *mountOptions.umaskString)
+		return false
+	}
+
+	if len(args) > 0 {
+		return false
+	}
+
+	return RunMount2(&mount2Options, os.FileMode(umask))
+}
+
+func RunMount2(option *Mount2Options, umask os.FileMode) bool {
+
+	// basic checks
+	chunkSizeLimitMB := *mountOptions.chunkSizeLimitMB
+	if chunkSizeLimitMB <= 0 {
+		fmt.Printf("Please specify a reasonable buffer size.")
+		return false
+	}
+
+	// try to connect to filer
+	filerAddresses := pb.ServerAddresses(*option.filer).ToAddresses()
+	util.LoadConfiguration("security", false)
+	grpcDialOption := security.LoadClientTLS(util.GetViper(), "grpc.client")
+	var cipher bool
+	var err error
+	for i := 0; i < 10; i++ {
+		err = pb.WithOneOfGrpcFilerClients(false, filerAddresses, grpcDialOption, func(client filer_pb.SeaweedFilerClient) error {
+			resp, err := client.GetFilerConfiguration(context.Background(), &filer_pb.GetFilerConfigurationRequest{})
+			if err != nil {
+				return fmt.Errorf("get filer grpc address %v configuration: %v", filerAddresses, err)
+			}
+			cipher = resp.Cipher
+			return nil
+		})
+		if err != nil {
+			glog.V(0).Infof("failed to talk to filer %v: %v", filerAddresses, err)
+			glog.V(0).Infof("wait for %d seconds ...", i+1)
+			time.Sleep(time.Duration(i+1) * time.Second)
+		}
+	}
+	if err != nil {
+		glog.Errorf("failed to talk to filer %v: %v", filerAddresses, err)
+		return true
+	}
+
+	filerMountRootPath := *option.filerMountRootPath
+
+	// clean up mount point
+	dir := util.ResolvePath(*option.dir)
+	if dir == "" {
+		fmt.Printf("Please specify the mount directory via \"-dir\"")
+		return false
+	}
+
+	unmount.Unmount(dir)
+
+	// detect mount folder mode
+	if *option.dirAutoCreate {
+		os.MkdirAll(dir, os.FileMode(0777)&^umask)
+	}
+	fileInfo, err := os.Stat(dir)
+
+	// collect uid, gid
+	uid, gid := uint32(0), uint32(0)
+	mountMode := os.ModeDir | 0777
+	if err == nil {
+		mountMode = os.ModeDir | os.FileMode(0777)&^umask
+		uid, gid = util.GetFileUidGid(fileInfo)
+		fmt.Printf("mount point owner uid=%d gid=%d mode=%s\n", uid, gid, mountMode)
+	} else {
+		fmt.Printf("can not stat %s\n", dir)
+		return false
+	}
+
+	// detect uid, gid
+	if uid == 0 {
+		if u, err := user.Current(); err == nil {
+			if parsedId, pe := strconv.ParseUint(u.Uid, 10, 32); pe == nil {
+				uid = uint32(parsedId)
+			}
+			if parsedId, pe := strconv.ParseUint(u.Gid, 10, 32); pe == nil {
+				gid = uint32(parsedId)
+			}
+			fmt.Printf("current uid=%d gid=%d\n", uid, gid)
+		}
+	}
+
+	// mapping uid, gid
+	uidGidMapper, err := meta_cache.NewUidGidMapper(*option.uidMap, *option.gidMap)
+	if err != nil {
+		fmt.Printf("failed to parse %s %s: %v\n", *option.uidMap, *option.gidMap, err)
+		return false
+	}
+
+	// Ensure target mount point availability
+	if isValid := checkMountPointAvailable(dir); !isValid {
+		glog.Fatalf("Expected mount to still be active, target mount point: %s, please check!", dir)
+		return true
+	}
+
+	// mount fuse
+	fuseMountOptions := &fuse.MountOptions{
+		AllowOther:               *option.allowOthers,
+		Options:                  nil,
+		MaxBackground:            128,
+		MaxWrite:                 1024 * 1024 * 2,
+		MaxReadAhead:             1024 * 1024 * 2,
+		IgnoreSecurityLabels:     false,
+		RememberInodes:           false,
+		FsName:                   *option.filer + ":" + filerMountRootPath,
+		Name:                     "seaweedfs",
+		SingleThreaded:           false,
+		DisableXAttrs:            false,
+		Debug:                    true,
+		EnableLocks:              false,
+		ExplicitDataCacheControl: false,
+		// SyncRead:                 false, // set to false to enable the FUSE_CAP_ASYNC_READ capability
+		DirectMount:      true,
+		DirectMountFlags: 0,
+		// EnableAcl:                false,
+	}
+
+	// find mount point
+	mountRoot := filerMountRootPath
+	if mountRoot != "/" && strings.HasSuffix(mountRoot, "/") {
+		mountRoot = mountRoot[0 : len(mountRoot)-1]
+	}
+
+	seaweedFileSystem := mount.NewSeaweedFileSystem(&mount.Option{
+		MountDirectory:     dir,
+		FilerAddresses:     filerAddresses,
+		GrpcDialOption:     grpcDialOption,
+		FilerMountRootPath: mountRoot,
+		Collection:         *option.collection,
+		Replication:        *option.replication,
+		TtlSec:             int32(*option.ttlSec),
+		DiskType:           types.ToDiskType(*option.diskType),
+		ChunkSizeLimit:     int64(chunkSizeLimitMB) * 1024 * 1024,
+		ConcurrentWriters:  *option.concurrentWriters,
+		CacheDir:           *option.cacheDir,
+		CacheSizeMB:        *option.cacheSizeMB,
+		DataCenter:         *option.dataCenter,
+		MountUid:           uid,
+		MountGid:           gid,
+		MountMode:          mountMode,
+		MountCtime:         fileInfo.ModTime(),
+		MountMtime:         time.Now(),
+		Umask:              umask,
+		VolumeServerAccess: *mountOptions.volumeServerAccess,
+		Cipher:             cipher,
+		UidGidMapper:       uidGidMapper,
+	})
+
+	server, err := fuse.NewServer(seaweedFileSystem, dir, fuseMountOptions)
+	if err != nil {
+		glog.Fatalf("Mount fail: %v", err)
+	}
+	grace.OnInterrupt(func() {
+		unmount.Unmount(dir)
+	})
+
+	seaweedFileSystem.StartBackgroundTasks()
+
+	fmt.Printf("This is SeaweedFS version %s %s %s\n", util.Version(), runtime.GOOS, runtime.GOARCH)
+
+	server.Serve()
+
+	return true
+}

weed/command/scaffold.go

@@ -2,7 +2,7 @@ package command
 
 import (
 	"fmt"
-	"os"
+	"github.com/chrislusf/seaweedfs/weed/util"
 	"path/filepath"
 
 	"github.com/chrislusf/seaweedfs/weed/command/scaffold"
@@ -56,7 +56,7 @@ func runScaffold(cmd *Command, args []string) bool {
 	}
 
 	if *outputPath != "" {
-		os.WriteFile(filepath.Join(*outputPath, *config+".toml"), []byte(content), 0644)
+		util.WriteFile(filepath.Join(*outputPath, *config+".toml"), []byte(content), 0644)
 	} else {
 		fmt.Println(content)
 	}

weed/filer/entry.go

@@ -43,6 +43,7 @@ type Entry struct {
 	HardLinkCounter int32
 	Content         []byte
 	Remote          *filer_pb.RemoteEntry
+	Quota           int64
 }
 
 func (entry *Entry) Size() uint64 {
@@ -70,6 +71,7 @@ func (entry *Entry) ShallowClone() *Entry {
 	newEntry.HardLinkCounter = entry.HardLinkCounter
 	newEntry.Content = entry.Content
 	newEntry.Remote = entry.Remote
+	newEntry.Quota = entry.Quota
 
 	return newEntry
 }
@@ -96,6 +98,7 @@ func (entry *Entry) ToExistingProtoEntry(message *filer_pb.Entry) {
 	message.HardLinkCounter = entry.HardLinkCounter
 	message.Content = entry.Content
 	message.RemoteEntry = entry.Remote
+	message.Quota = entry.Quota
 }
 
 func FromPbEntryToExistingEntry(message *filer_pb.Entry, fsEntry *Entry) {
@@ -106,6 +109,7 @@ func FromPbEntryToExistingEntry(message *filer_pb.Entry, fsEntry *Entry) {
 	fsEntry.HardLinkCounter = message.HardLinkCounter
 	fsEntry.Content = message.Content
 	fsEntry.Remote = message.RemoteEntry
+	fsEntry.Quota = message.Quota
 }
 
 func (entry *Entry) ToProtoFullEntry() *filer_pb.FullEntry {

weed/filer/entry_codec.go

@@ -118,6 +118,9 @@ func EqualEntry(a, b *Entry) bool {
 	if !proto.Equal(a.Remote, b.Remote) {
 		return false
 	}
+	if a.Quota != b.Quota {
+		return false
+	}
 	return true
 }
 

weed/filer/filechunks.go

@@ -101,6 +101,21 @@ func DoMinusChunks(as, bs []*filer_pb.FileChunk) (delta []*filer_pb.FileChunk) {
 	return
 }
 
+func DoMinusChunksBySourceFileId(as, bs []*filer_pb.FileChunk) (delta []*filer_pb.FileChunk) {
+
+	fileIds := make(map[string]bool)
+	for _, interval := range bs {
+		fileIds[interval.GetFileIdString()] = true
+	}
+	for _, chunk := range as {
+		if _, found := fileIds[chunk.GetSourceFileId()]; !found {
+			delta = append(delta, chunk)
+		}
+	}
+
+	return
+}
+
 type ChunkView struct {
 	FileId      string
 	Offset      int64

weed/filer/filer_conf.go

@@ -75,12 +75,12 @@ func (fc *FilerConf) loadFromFiler(filer *Filer) (err error) {
 		return fc.LoadFromBytes(entry.Content)
 	}
 
-	return fc.loadFromChunks(filer, entry.Content, entry.Chunks)
+	return fc.loadFromChunks(filer, entry.Content, entry.Chunks, entry.Size())
 }
 
-func (fc *FilerConf) loadFromChunks(filer *Filer, content []byte, chunks []*filer_pb.FileChunk) (err error) {
+func (fc *FilerConf) loadFromChunks(filer *Filer, content []byte, chunks []*filer_pb.FileChunk, size uint64) (err error) {
 	if len(content) == 0 {
-		content, err = filer.readEntry(chunks)
+		content, err = filer.readEntry(chunks, size)
 		if err != nil {
 			glog.Errorf("read filer conf content: %v", err)
 			return

weed/filer/filer_on_meta_event.go

@@ -2,8 +2,6 @@ package filer
 
 import (
 	"bytes"
-	"math"
-
 	"github.com/chrislusf/seaweedfs/weed/glog"
 	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
 	"github.com/chrislusf/seaweedfs/weed/util"
@@ -55,9 +53,9 @@ func (f *Filer) maybeReloadFilerConfiguration(event *filer_pb.SubscribeMetadataR
 	}
 }
 
-func (f *Filer) readEntry(chunks []*filer_pb.FileChunk) ([]byte, error) {
+func (f *Filer) readEntry(chunks []*filer_pb.FileChunk, size uint64) ([]byte, error) {
 	var buf bytes.Buffer
-	err := StreamContent(f.MasterClient, &buf, chunks, 0, math.MaxInt64)
+	err := StreamContent(f.MasterClient, &buf, chunks, 0, int64(size))
 	if err != nil {
 		return nil, err
 	}
@@ -66,7 +64,7 @@ func (f *Filer) readEntry(chunks []*filer_pb.FileChunk) ([]byte, error) {
 
 func (f *Filer) reloadFilerConfiguration(entry *filer_pb.Entry) {
 	fc := NewFilerConf()
-	err := fc.loadFromChunks(f, entry.Content, entry.Chunks)
+	err := fc.loadFromChunks(f, entry.Content, entry.Chunks, FileSize(entry))
 	if err != nil {
 		glog.Errorf("read filer conf chunks: %v", err)
 		return

weed/filer/meta_aggregator.go

@@ -165,13 +165,14 @@ func (ma *MetaAggregator) subscribeToOneFiler(f *Filer, self pb.ServerAddress, p
 
 	for {
 		glog.V(4).Infof("subscribing remote %s meta change: %v", peer, time.Unix(0, lastTsNs))
-		err := pb.WithFilerClient(false, peer, ma.grpcDialOption, func(client filer_pb.SeaweedFilerClient) error {
+		err := pb.WithFilerClient(true, peer, ma.grpcDialOption, func(client filer_pb.SeaweedFilerClient) error {
 			ctx, cancel := context.WithCancel(context.Background())
 			defer cancel()
 			stream, err := client.SubscribeLocalMetadata(ctx, &filer_pb.SubscribeMetadataRequest{
 				ClientName: "filer:" + string(self),
 				PathPrefix: "/",
 				SinceNs:    lastTsNs,
+				ClientId:   int32(ma.filer.UniqueFileId),
 			})
 			if err != nil {
 				return fmt.Errorf("subscribe: %v", err)

weed/filer/read_write.go

@@ -4,7 +4,6 @@ import (
 	"bytes"
 	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
 	"github.com/chrislusf/seaweedfs/weed/wdclient"
-	"math"
 	"time"
 )
 
@@ -23,7 +22,7 @@ func ReadEntry(masterClient *wdclient.MasterClient, filerClient filer_pb.Seaweed
 		return err
 	}
 
-	return StreamContent(masterClient, byteBuffer, respLookupEntry.Entry.Chunks, 0, math.MaxInt64)
+	return StreamContent(masterClient, byteBuffer, respLookupEntry.Entry.Chunks, 0, int64(FileSize(respLookupEntry.Entry)))
 
 }
 

weed/filesys/dir.go

@@ -491,8 +491,8 @@ func (dir *Dir) removeFolder(entry *filer_pb.Entry, req *fuse.RemoveRequest) err
 	dirFullPath := dir.FullPath()
 
 	glog.V(3).Infof("remove directory entry: %v", req)
-	ignoreRecursiveErr := false // ignore recursion error since the OS should manage it
-	err := filer_pb.Remove(dir.wfs, dirFullPath, req.Name, true, false, ignoreRecursiveErr, false, []int32{dir.wfs.signature})
+	ignoreRecursiveErr := true // ignore recursion error since the OS should manage it
+	err := filer_pb.Remove(dir.wfs, dirFullPath, req.Name, true, true, ignoreRecursiveErr, false, []int32{dir.wfs.signature})
 	if err != nil {
 		glog.V(0).Infof("remove %s/%s: %v", dirFullPath, req.Name, err)
 		if strings.Contains(err.Error(), filer.MsgFailDelNonEmptyFolder) {

weed/filesys/dirty_pages_chunked.go

@@ -10,7 +10,7 @@ import (
 	"time"
 )
 
-type MemoryChunkPages struct {
+type ChunkedDirtyPages struct {
 	fh             *FileHandle
 	writeWaitGroup sync.WaitGroup
 	chunkAddLock   sync.Mutex
@@ -22,22 +22,21 @@ type MemoryChunkPages struct {
 }
 
 var (
-	_ = page_writer.DirtyPages(&MemoryChunkPages{})
+	_ = page_writer.DirtyPages(&ChunkedDirtyPages{})
 )
 
-func newMemoryChunkPages(fh *FileHandle, chunkSize int64) *MemoryChunkPages {
+func newMemoryChunkPages(fh *FileHandle, chunkSize int64) *ChunkedDirtyPages {
 
-	dirtyPages := &MemoryChunkPages{
+	dirtyPages := &ChunkedDirtyPages{
 		fh: fh,
 	}
 
-	dirtyPages.uploadPipeline = page_writer.NewUploadPipeline(fh.f.fullpath(),
-		fh.f.wfs.concurrentWriters, chunkSize, dirtyPages.saveChunkedFileIntevalToStorage)
+	dirtyPages.uploadPipeline = page_writer.NewUploadPipeline(fh.f.wfs.concurrentWriters, chunkSize, dirtyPages.saveChunkedFileIntevalToStorage, fh.f.wfs.option.ConcurrentWriters)
 
 	return dirtyPages
 }
 
-func (pages *MemoryChunkPages) AddPage(offset int64, data []byte) {
+func (pages *ChunkedDirtyPages) AddPage(offset int64, data []byte) {
 	pages.hasWrites = true
 
 	glog.V(4).Infof("%v memory AddPage [%d, %d)", pages.fh.f.fullpath(), offset, offset+int64(len(data)))
@@ -46,7 +45,7 @@ func (pages *MemoryChunkPages) AddPage(offset int64, data []byte) {
 	return
 }
 
-func (pages *MemoryChunkPages) FlushData() error {
+func (pages *ChunkedDirtyPages) FlushData() error {
 	if !pages.hasWrites {
 		return nil
 	}
@@ -57,18 +56,18 @@ func (pages *MemoryChunkPages) FlushData() error {
 	return nil
 }
 
-func (pages *MemoryChunkPages) ReadDirtyDataAt(data []byte, startOffset int64) (maxStop int64) {
+func (pages *ChunkedDirtyPages) ReadDirtyDataAt(data []byte, startOffset int64) (maxStop int64) {
 	if !pages.hasWrites {
 		return
 	}
 	return pages.uploadPipeline.MaybeReadDataAt(data, startOffset)
 }
 
-func (pages *MemoryChunkPages) GetStorageOptions() (collection, replication string) {
+func (pages *ChunkedDirtyPages) GetStorageOptions() (collection, replication string) {
 	return pages.collection, pages.replication
 }
 
-func (pages *MemoryChunkPages) saveChunkedFileIntevalToStorage(reader io.Reader, offset int64, size int64, cleanupFn func()) {
+func (pages *ChunkedDirtyPages) saveChunkedFileIntevalToStorage(reader io.Reader, offset int64, size int64, cleanupFn func()) {
 
 	mtime := time.Now().UnixNano()
 	defer cleanupFn()
@@ -89,13 +88,13 @@ func (pages *MemoryChunkPages) saveChunkedFileIntevalToStorage(reader io.Reader,
 
 }
 
-func (pages MemoryChunkPages) Destroy() {
+func (pages ChunkedDirtyPages) Destroy() {
 	pages.uploadPipeline.Shutdown()
 }
 
-func (pages *MemoryChunkPages) LockForRead(startOffset, stopOffset int64) {
+func (pages *ChunkedDirtyPages) LockForRead(startOffset, stopOffset int64) {
 	pages.uploadPipeline.LockForRead(startOffset, stopOffset)
 }
-func (pages *MemoryChunkPages) UnlockForRead(startOffset, stopOffset int64) {
+func (pages *ChunkedDirtyPages) UnlockForRead(startOffset, stopOffset int64) {
 	pages.uploadPipeline.UnlockForRead(startOffset, stopOffset)
 }

weed/filesys/dirty_pages_temp_file.go

@@ -1,112 +0,0 @@
-package filesys
-
-import (
-	"fmt"
-	"github.com/chrislusf/seaweedfs/weed/filesys/page_writer"
-	"github.com/chrislusf/seaweedfs/weed/glog"
-	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
-	"io"
-	"os"
-	"sync"
-	"time"
-)
-
-type TempFileDirtyPages struct {
-	f              *File
-	writeWaitGroup sync.WaitGroup
-	pageAddLock    sync.Mutex
-	chunkAddLock   sync.Mutex
-	lastErr        error
-	collection     string
-	replication    string
-	chunkedFile    *page_writer.ChunkedFileWriter
-}
-
-func newTempFileDirtyPages(file *File, chunkSize int64) *TempFileDirtyPages {
-
-	tempFile := &TempFileDirtyPages{
-		f:           file,
-		chunkedFile: page_writer.NewChunkedFileWriter(file.wfs.option.getTempFilePageDir(), chunkSize),
-	}
-
-	return tempFile
-}
-
-func (pages *TempFileDirtyPages) AddPage(offset int64, data []byte) {
-
-	pages.pageAddLock.Lock()
-	defer pages.pageAddLock.Unlock()
-
-	glog.V(4).Infof("%v tempfile AddPage [%d, %d)", pages.f.fullpath(), offset, offset+int64(len(data)))
-	if _, err := pages.chunkedFile.WriteAt(data, offset); err != nil {
-		pages.lastErr = err
-	}
-
-	return
-}
-
-func (pages *TempFileDirtyPages) FlushData() error {
-	pages.saveChunkedFileToStorage()
-	pages.writeWaitGroup.Wait()
-	if pages.lastErr != nil {
-		return fmt.Errorf("flush data: %v", pages.lastErr)
-	}
-	pages.chunkedFile.Reset()
-	return nil
-}
-
-func (pages *TempFileDirtyPages) ReadDirtyDataAt(data []byte, startOffset int64) (maxStop int64) {
-	return pages.chunkedFile.ReadDataAt(data, startOffset)
-}
-
-func (pages *TempFileDirtyPages) GetStorageOptions() (collection, replication string) {
-	return pages.collection, pages.replication
-}
-
-func (pages *TempFileDirtyPages) saveChunkedFileToStorage() {
-
-	pages.chunkedFile.ProcessEachInterval(func(file *os.File, logicChunkIndex page_writer.LogicChunkIndex, interval *page_writer.ChunkWrittenInterval) {
-		reader := page_writer.NewFileIntervalReader(pages.chunkedFile, logicChunkIndex, interval)
-		pages.saveChunkedFileIntevalToStorage(reader, int64(logicChunkIndex)*pages.chunkedFile.ChunkSize+interval.StartOffset, interval.Size())
-	})
-
-}
-
-func (pages *TempFileDirtyPages) saveChunkedFileIntevalToStorage(reader io.Reader, offset int64, size int64) {
-
-	mtime := time.Now().UnixNano()
-	pages.writeWaitGroup.Add(1)
-	writer := func() {
-		defer pages.writeWaitGroup.Done()
-
-		chunk, collection, replication, err := pages.f.wfs.saveDataAsChunk(pages.f.fullpath())(reader, pages.f.Name, offset)
-		if err != nil {
-			glog.V(0).Infof("%s saveToStorage [%d,%d): %v", pages.f.fullpath(), offset, offset+size, err)
-			pages.lastErr = err
-			return
-		}
-		chunk.Mtime = mtime
-		pages.collection, pages.replication = collection, replication
-		pages.chunkAddLock.Lock()
-		defer pages.chunkAddLock.Unlock()
-		pages.f.addChunks([]*filer_pb.FileChunk{chunk})
-		glog.V(3).Infof("%s saveToStorage %s [%d,%d)", pages.f.fullpath(), chunk.FileId, offset, offset+size)
-	}
-
-	if pages.f.wfs.concurrentWriters != nil {
-		pages.f.wfs.concurrentWriters.Execute(writer)
-	} else {
-		go writer()
-	}
-
-}
-
-func (pages TempFileDirtyPages) Destroy() {
-	pages.chunkedFile.Reset()
-}
-
-func (pages *TempFileDirtyPages) LockForRead(startOffset, stopOffset int64) {
-}
-
-func (pages *TempFileDirtyPages) UnlockForRead(startOffset, stopOffset int64) {
-}

weed/filesys/file.go

@@ -72,8 +72,8 @@ func (file *File) Attr(ctx context.Context, attr *fuse.Attr) (err error) {
 	attr.Mtime = time.Unix(entry.Attributes.Mtime, 0)
 	attr.Gid = entry.Attributes.Gid
 	attr.Uid = entry.Attributes.Uid
-	attr.Blocks = attr.Size/blockSize + 1
-	attr.BlockSize = uint32(file.wfs.option.ChunkSizeLimit)
+	attr.BlockSize = blockSize
+	attr.Blocks = (attr.Size + blockSize - 1) / blockSize
 	if entry.HardLinkCounter > 0 {
 		attr.Nlink = uint32(entry.HardLinkCounter)
 	}

weed/filesys/filehandle.go

@@ -158,7 +158,7 @@ func (fh *FileHandle) readFromChunks(buff []byte, offset int64) (int64, error) {
 	reader := fh.reader
 	if reader == nil {
 		chunkViews := filer.ViewFromVisibleIntervals(fh.entryViewCache, 0, math.MaxInt64)
-		glog.V(4).Infof("file handle read %s [%d,%d) from %+v", fileFullPath, offset, offset+int64(len(buff)), chunkViews)
+		glog.V(4).Infof("file handle read %s [%d,%d) from %d views", fileFullPath, offset, offset+int64(len(buff)), len(chunkViews))
 		for _, chunkView := range chunkViews {
 			glog.V(4).Infof("  read %s [%d,%d) from chunk %+v", fileFullPath, chunkView.LogicOffset, chunkView.LogicOffset+int64(chunkView.Size), chunkView.FileId)
 		}

weed/filesys/page_writer/chunk_interval_list.go

@@ -51,6 +51,12 @@ func (list *ChunkWrittenIntervalList) MarkWritten(startOffset, stopOffset int64)
 func (list *ChunkWrittenIntervalList) IsComplete(chunkSize int64) bool {
 	return list.size() == 1 && list.head.next.isComplete(chunkSize)
 }
+func (list *ChunkWrittenIntervalList) WrittenSize() (writtenByteCount int64) {
+	for t := list.head; t != nil; t = t.next {
+		writtenByteCount += t.Size()
+	}
+	return
+}
 
 func (list *ChunkWrittenIntervalList) addInterval(interval *ChunkWrittenInterval) {
 

weed/filesys/page_writer/chunked_file_writer.go

@@ -1,159 +0,0 @@
-package page_writer
-
-import (
-	"github.com/chrislusf/seaweedfs/weed/glog"
-	"io"
-	"os"
-	"sync"
-)
-
-type ActualChunkIndex int
-
-// ChunkedFileWriter assumes the write requests will come in within chunks
-type ChunkedFileWriter struct {
-	dir                     string
-	file                    *os.File
-	logicToActualChunkIndex map[LogicChunkIndex]ActualChunkIndex
-	chunkUsages             []*ChunkWrittenIntervalList
-	ChunkSize               int64
-	sync.Mutex
-}
-
-var _ = io.WriterAt(&ChunkedFileWriter{})
-
-func NewChunkedFileWriter(dir string, chunkSize int64) *ChunkedFileWriter {
-	return &ChunkedFileWriter{
-		dir:                     dir,
-		file:                    nil,
-		logicToActualChunkIndex: make(map[LogicChunkIndex]ActualChunkIndex),
-		ChunkSize:               chunkSize,
-	}
-}
-
-func (cw *ChunkedFileWriter) WriteAt(p []byte, off int64) (n int, err error) {
-	cw.Lock()
-	defer cw.Unlock()
-
-	if cw.file == nil {
-		cw.file, err = os.CreateTemp(cw.dir, "")
-		if err != nil {
-			glog.Errorf("create temp file: %v", err)
-			return
-		}
-	}
-
-	actualOffset, chunkUsage := cw.toActualWriteOffset(off)
-	n, err = cw.file.WriteAt(p, actualOffset)
-	if err == nil {
-		startOffset := off % cw.ChunkSize
-		chunkUsage.MarkWritten(startOffset, startOffset+int64(n))
-	}
-	return
-}
-
-func (cw *ChunkedFileWriter) ReadDataAt(p []byte, off int64) (maxStop int64) {
-	cw.Lock()
-	defer cw.Unlock()
-
-	if cw.file == nil {
-		return
-	}
-
-	logicChunkIndex := off / cw.ChunkSize
-	actualChunkIndex, chunkUsage := cw.toActualReadOffset(off)
-	if chunkUsage != nil {
-		for t := chunkUsage.head.next; t != chunkUsage.tail; t = t.next {
-			logicStart := max(off, logicChunkIndex*cw.ChunkSize+t.StartOffset)
-			logicStop := min(off+int64(len(p)), logicChunkIndex*cw.ChunkSize+t.stopOffset)
-			if logicStart < logicStop {
-				actualStart := logicStart - logicChunkIndex*cw.ChunkSize + int64(actualChunkIndex)*cw.ChunkSize
-				_, err := cw.file.ReadAt(p[logicStart-off:logicStop-off], actualStart)
-				if err != nil {
-					glog.Errorf("reading temp file: %v", err)
-					break
-				}
-				maxStop = max(maxStop, logicStop)
-			}
-		}
-	}
-	return
-}
-
-func (cw *ChunkedFileWriter) toActualWriteOffset(logicOffset int64) (actualOffset int64, chunkUsage *ChunkWrittenIntervalList) {
-	logicChunkIndex := LogicChunkIndex(logicOffset / cw.ChunkSize)
-	offsetRemainder := logicOffset % cw.ChunkSize
-	existingActualChunkIndex, found := cw.logicToActualChunkIndex[logicChunkIndex]
-	if found {
-		return int64(existingActualChunkIndex)*cw.ChunkSize + offsetRemainder, cw.chunkUsages[existingActualChunkIndex]
-	}
-	cw.logicToActualChunkIndex[logicChunkIndex] = ActualChunkIndex(len(cw.chunkUsages))
-	chunkUsage = newChunkWrittenIntervalList()
-	cw.chunkUsages = append(cw.chunkUsages, chunkUsage)
-	return int64(len(cw.chunkUsages)-1)*cw.ChunkSize + offsetRemainder, chunkUsage
-}
-
-func (cw *ChunkedFileWriter) toActualReadOffset(logicOffset int64) (actualChunkIndex ActualChunkIndex, chunkUsage *ChunkWrittenIntervalList) {
-	logicChunkIndex := LogicChunkIndex(logicOffset / cw.ChunkSize)
-	existingActualChunkIndex, found := cw.logicToActualChunkIndex[logicChunkIndex]
-	if found {
-		return existingActualChunkIndex, cw.chunkUsages[existingActualChunkIndex]
-	}
-	return 0, nil
-}
-
-func (cw *ChunkedFileWriter) ProcessEachInterval(process func(file *os.File, logicChunkIndex LogicChunkIndex, interval *ChunkWrittenInterval)) {
-	for logicChunkIndex, actualChunkIndex := range cw.logicToActualChunkIndex {
-		chunkUsage := cw.chunkUsages[actualChunkIndex]
-		for t := chunkUsage.head.next; t != chunkUsage.tail; t = t.next {
-			process(cw.file, logicChunkIndex, t)
-		}
-	}
-}
-
-// Reset releases used resources
-func (cw *ChunkedFileWriter) Reset() {
-	if cw.file != nil {
-		cw.file.Close()
-		os.Remove(cw.file.Name())
-		cw.file = nil
-	}
-	cw.logicToActualChunkIndex = make(map[LogicChunkIndex]ActualChunkIndex)
-	cw.chunkUsages = cw.chunkUsages[:0]
-}
-
-type FileIntervalReader struct {
-	f           *os.File
-	startOffset int64
-	stopOffset  int64
-	position    int64
-}
-
-var _ = io.Reader(&FileIntervalReader{})
-
-func NewFileIntervalReader(cw *ChunkedFileWriter, logicChunkIndex LogicChunkIndex, interval *ChunkWrittenInterval) *FileIntervalReader {
-	actualChunkIndex, found := cw.logicToActualChunkIndex[logicChunkIndex]
-	if !found {
-		// this should never happen
-		return nil
-	}
-	return &FileIntervalReader{
-		f:           cw.file,
-		startOffset: int64(actualChunkIndex)*cw.ChunkSize + interval.StartOffset,
-		stopOffset:  int64(actualChunkIndex)*cw.ChunkSize + interval.stopOffset,
-		position:    0,
-	}
-}
-
-func (fr *FileIntervalReader) Read(p []byte) (n int, err error) {
-	readSize := minInt(len(p), int(fr.stopOffset-fr.startOffset-fr.position))
-	n, err = fr.f.ReadAt(p[:readSize], fr.startOffset+fr.position)
-	if err == nil || err == io.EOF {
-		fr.position += int64(n)
-		if fr.stopOffset-fr.startOffset-fr.position == 0 {
-			// return a tiny bit faster
-			err = io.EOF
-			return
-		}
-	}
-	return
-}

weed/filesys/page_writer/chunked_file_writer_test.go

@@ -1,60 +0,0 @@
-package page_writer
-
-import (
-	"github.com/stretchr/testify/assert"
-	"os"
-	"testing"
-)
-
-func TestChunkedFileWriter_toActualOffset(t *testing.T) {
-	cw := NewChunkedFileWriter("", 16)
-
-	writeToFile(cw, 50, 60)
-	writeToFile(cw, 60, 64)
-
-	writeToFile(cw, 32, 40)
-	writeToFile(cw, 42, 48)
-
-	writeToFile(cw, 48, 50)
-
-	assert.Equal(t, 1, cw.chunkUsages[0].size(), "fully covered")
-	assert.Equal(t, 2, cw.chunkUsages[1].size(), "2 intervals")
-
-}
-
-func writeToFile(cw *ChunkedFileWriter, startOffset int64, stopOffset int64) {
-
-	_, chunkUsage := cw.toActualWriteOffset(startOffset)
-
-	// skip doing actual writing
-
-	innerOffset := startOffset % cw.ChunkSize
-	chunkUsage.MarkWritten(innerOffset, innerOffset+stopOffset-startOffset)
-
-}
-
-func TestWriteChunkedFile(t *testing.T) {
-	x := NewChunkedFileWriter(os.TempDir(), 20)
-	defer x.Reset()
-	y := NewChunkedFileWriter(os.TempDir(), 12)
-	defer y.Reset()
-
-	batchSize := 4
-	buf := make([]byte, batchSize)
-	for i := 0; i < 256; i++ {
-		for x := 0; x < batchSize; x++ {
-			buf[x] = byte(i)
-		}
-		x.WriteAt(buf, int64(i*batchSize))
-		y.WriteAt(buf, int64((255-i)*batchSize))
-	}
-
-	a := make([]byte, 1)
-	b := make([]byte, 1)
-	for i := 0; i < 256*batchSize; i++ {
-		x.ReadDataAt(a, int64(i))
-		y.ReadDataAt(b, int64(256*batchSize-1-i))
-		assert.Equal(t, a[0], b[0], "same read")
-	}
-
-}

weed/filesys/page_writer/chunked_stream_writer.go

@@ -1,12 +0,0 @@
-package page_writer
-
-import (
-	"io"
-)
-
-type SaveToStorageFunc func(reader io.Reader, offset int64, size int64, cleanupFn func())
-
-type MemChunk struct {
-	buf   []byte
-	usage *ChunkWrittenIntervalList
-}

weed/filesys/page_writer/page_chunk.go

@@ -0,0 +1,16 @@
+package page_writer
+
+import (
+	"io"
+)
+
+type SaveToStorageFunc func(reader io.Reader, offset int64, size int64, cleanupFn func())
+
+type PageChunk interface {
+	FreeResource()
+	WriteDataAt(src []byte, offset int64) (n int)
+	ReadDataAt(p []byte, off int64) (maxStop int64)
+	IsComplete() bool
+	WrittenSize() int64
+	SaveContent(saveFn SaveToStorageFunc)
+}

weed/filesys/page_writer/page_chunk_mem.go

@@ -0,0 +1,69 @@
+package page_writer
+
+import (
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"github.com/chrislusf/seaweedfs/weed/util/mem"
+)
+
+var (
+	_ = PageChunk(&MemChunk{})
+)
+
+type MemChunk struct {
+	buf             []byte
+	usage           *ChunkWrittenIntervalList
+	chunkSize       int64
+	logicChunkIndex LogicChunkIndex
+}
+
+func NewMemChunk(logicChunkIndex LogicChunkIndex, chunkSize int64) *MemChunk {
+	return &MemChunk{
+		logicChunkIndex: logicChunkIndex,
+		chunkSize:       chunkSize,
+		buf:             mem.Allocate(int(chunkSize)),
+		usage:           newChunkWrittenIntervalList(),
+	}
+}
+
+func (mc *MemChunk) FreeResource() {
+	mem.Free(mc.buf)
+}
+
+func (mc *MemChunk) WriteDataAt(src []byte, offset int64) (n int) {
+	innerOffset := offset % mc.chunkSize
+	n = copy(mc.buf[innerOffset:], src)
+	mc.usage.MarkWritten(innerOffset, innerOffset+int64(n))
+	return
+}
+
+func (mc *MemChunk) ReadDataAt(p []byte, off int64) (maxStop int64) {
+	memChunkBaseOffset := int64(mc.logicChunkIndex) * mc.chunkSize
+	for t := mc.usage.head.next; t != mc.usage.tail; t = t.next {
+		logicStart := max(off, int64(mc.logicChunkIndex)*mc.chunkSize+t.StartOffset)
+		logicStop := min(off+int64(len(p)), memChunkBaseOffset+t.stopOffset)
+		if logicStart < logicStop {
+			copy(p[logicStart-off:logicStop-off], mc.buf[logicStart-memChunkBaseOffset:logicStop-memChunkBaseOffset])
+			maxStop = max(maxStop, logicStop)
+		}
+	}
+	return
+}
+
+func (mc *MemChunk) IsComplete() bool {
+	return mc.usage.IsComplete(mc.chunkSize)
+}
+
+func (mc *MemChunk) WrittenSize() int64 {
+	return mc.usage.WrittenSize()
+}
+
+func (mc *MemChunk) SaveContent(saveFn SaveToStorageFunc) {
+	if saveFn == nil {
+		return
+	}
+	for t := mc.usage.head.next; t != mc.usage.tail; t = t.next {
+		reader := util.NewBytesReader(mc.buf[t.StartOffset:t.stopOffset])
+		saveFn(reader, int64(mc.logicChunkIndex)*mc.chunkSize+t.StartOffset, t.Size(), func() {
+		})
+	}
+}

weed/filesys/page_writer/page_chunk_swapfile.go

@@ -0,0 +1,121 @@
+package page_writer
+
+import (
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"github.com/chrislusf/seaweedfs/weed/util/mem"
+	"os"
+)
+
+var (
+	_ = PageChunk(&SwapFileChunk{})
+)
+
+type ActualChunkIndex int
+
+type SwapFile struct {
+	dir                     string
+	file                    *os.File
+	logicToActualChunkIndex map[LogicChunkIndex]ActualChunkIndex
+	chunkSize               int64
+}
+
+type SwapFileChunk struct {
+	swapfile         *SwapFile
+	usage            *ChunkWrittenIntervalList
+	logicChunkIndex  LogicChunkIndex
+	actualChunkIndex ActualChunkIndex
+}
+
+func NewSwapFile(dir string, chunkSize int64) *SwapFile {
+	return &SwapFile{
+		dir:                     dir,
+		file:                    nil,
+		logicToActualChunkIndex: make(map[LogicChunkIndex]ActualChunkIndex),
+		chunkSize:               chunkSize,
+	}
+}
+func (sf *SwapFile) FreeResource() {
+	if sf.file != nil {
+		sf.file.Close()
+		os.Remove(sf.file.Name())
+	}
+}
+
+func (sf *SwapFile) NewTempFileChunk(logicChunkIndex LogicChunkIndex) (tc *SwapFileChunk) {
+	if sf.file == nil {
+		var err error
+		sf.file, err = os.CreateTemp(sf.dir, "")
+		if err != nil {
+			glog.Errorf("create swap file: %v", err)
+			return nil
+		}
+	}
+	actualChunkIndex, found := sf.logicToActualChunkIndex[logicChunkIndex]
+	if !found {
+		actualChunkIndex = ActualChunkIndex(len(sf.logicToActualChunkIndex))
+		sf.logicToActualChunkIndex[logicChunkIndex] = actualChunkIndex
+	}
+
+	return &SwapFileChunk{
+		swapfile:         sf,
+		usage:            newChunkWrittenIntervalList(),
+		logicChunkIndex:  logicChunkIndex,
+		actualChunkIndex: actualChunkIndex,
+	}
+}
+
+func (sc *SwapFileChunk) FreeResource() {
+}
+
+func (sc *SwapFileChunk) WriteDataAt(src []byte, offset int64) (n int) {
+	innerOffset := offset % sc.swapfile.chunkSize
+	var err error
+	n, err = sc.swapfile.file.WriteAt(src, int64(sc.actualChunkIndex)*sc.swapfile.chunkSize+innerOffset)
+	if err == nil {
+		sc.usage.MarkWritten(innerOffset, innerOffset+int64(n))
+	} else {
+		glog.Errorf("failed to write swap file %s: %v", sc.swapfile.file.Name(), err)
+	}
+	return
+}
+
+func (sc *SwapFileChunk) ReadDataAt(p []byte, off int64) (maxStop int64) {
+	chunkStartOffset := int64(sc.logicChunkIndex) * sc.swapfile.chunkSize
+	for t := sc.usage.head.next; t != sc.usage.tail; t = t.next {
+		logicStart := max(off, chunkStartOffset+t.StartOffset)
+		logicStop := min(off+int64(len(p)), chunkStartOffset+t.stopOffset)
+		if logicStart < logicStop {
+			actualStart := logicStart - chunkStartOffset + int64(sc.actualChunkIndex)*sc.swapfile.chunkSize
+			if _, err := sc.swapfile.file.ReadAt(p[logicStart-off:logicStop-off], actualStart); err != nil {
+				glog.Errorf("failed to reading swap file %s: %v", sc.swapfile.file.Name(), err)
+				break
+			}
+			maxStop = max(maxStop, logicStop)
+		}
+	}
+	return
+}
+
+func (sc *SwapFileChunk) IsComplete() bool {
+	return sc.usage.IsComplete(sc.swapfile.chunkSize)
+}
+
+func (sc *SwapFileChunk) WrittenSize() int64 {
+	return sc.usage.WrittenSize()
+}
+
+func (sc *SwapFileChunk) SaveContent(saveFn SaveToStorageFunc) {
+	if saveFn == nil {
+		return
+	}
+	for t := sc.usage.head.next; t != sc.usage.tail; t = t.next {
+		data := mem.Allocate(int(t.Size()))
+		sc.swapfile.file.ReadAt(data, t.StartOffset+int64(sc.actualChunkIndex)*sc.swapfile.chunkSize)
+		reader := util.NewBytesReader(data)
+		saveFn(reader, int64(sc.logicChunkIndex)*sc.swapfile.chunkSize+t.StartOffset, t.Size(), func() {
+		})
+		mem.Free(data)
+	}
+	sc.usage = newChunkWrittenIntervalList()
+}

weed/filesys/page_writer/upload_pipeline.go

@@ -4,7 +4,6 @@ import (
 	"fmt"
 	"github.com/chrislusf/seaweedfs/weed/glog"
 	"github.com/chrislusf/seaweedfs/weed/util"
-	"github.com/chrislusf/seaweedfs/weed/util/mem"
 	"sync"
 	"sync/atomic"
 	"time"
@@ -15,20 +14,21 @@ type LogicChunkIndex int
 type UploadPipeline struct {
 	filepath             util.FullPath
 	ChunkSize            int64
-	writers              *util.LimitedConcurrentExecutor
-	writableChunks       map[LogicChunkIndex]*MemChunk
+	writableChunks       map[LogicChunkIndex]PageChunk
 	writableChunksLock   sync.Mutex
 	sealedChunks         map[LogicChunkIndex]*SealedChunk
 	sealedChunksLock     sync.Mutex
-	activeWriterCond     *sync.Cond
-	activeWriterCount    int32
+	uploaders            *util.LimitedConcurrentExecutor
+	uploaderCount        int32
+	uploaderCountCond    *sync.Cond
+	saveToStorageFn      SaveToStorageFunc
 	activeReadChunks     map[LogicChunkIndex]int
 	activeReadChunksLock sync.Mutex
-	saveToStorageFn      SaveToStorageFunc
+	bufferChunkLimit     int
 }
 
 type SealedChunk struct {
-	chunk            *MemChunk
+	chunk            PageChunk
 	referenceCounter int // track uploading or reading processes
 }
 
@@ -36,221 +36,147 @@ func (sc *SealedChunk) FreeReference(messageOnFree string) {
 	sc.referenceCounter--
 	if sc.referenceCounter == 0 {
 		glog.V(4).Infof("Free sealed chunk: %s", messageOnFree)
-		mem.Free(sc.chunk.buf)
+		sc.chunk.FreeResource()
 	}
 }
 
-func NewUploadPipeline(filepath util.FullPath, writers *util.LimitedConcurrentExecutor, chunkSize int64, saveToStorageFn SaveToStorageFunc) *UploadPipeline {
+func NewUploadPipeline(writers *util.LimitedConcurrentExecutor, chunkSize int64, saveToStorageFn SaveToStorageFunc, bufferChunkLimit int) *UploadPipeline {
 	return &UploadPipeline{
-		ChunkSize:        chunkSize,
-		writableChunks:   make(map[LogicChunkIndex]*MemChunk),
-		sealedChunks:     make(map[LogicChunkIndex]*SealedChunk),
-		writers:          writers,
-		activeWriterCond: sync.NewCond(&sync.Mutex{}),
-		saveToStorageFn:  saveToStorageFn,
-		filepath:         filepath,
-		activeReadChunks: make(map[LogicChunkIndex]int),
+		ChunkSize:         chunkSize,
+		writableChunks:    make(map[LogicChunkIndex]PageChunk),
+		sealedChunks:      make(map[LogicChunkIndex]*SealedChunk),
+		uploaders:         writers,
+		uploaderCountCond: sync.NewCond(&sync.Mutex{}),
+		saveToStorageFn:   saveToStorageFn,
+		activeReadChunks:  make(map[LogicChunkIndex]int),
+		bufferChunkLimit:  bufferChunkLimit,
 	}
 }
 
-func (cw *UploadPipeline) SaveDataAt(p []byte, off int64) (n int) {
-	cw.writableChunksLock.Lock()
-	defer cw.writableChunksLock.Unlock()
+func (up *UploadPipeline) SaveDataAt(p []byte, off int64) (n int) {
+	up.writableChunksLock.Lock()
+	defer up.writableChunksLock.Unlock()
 
-	logicChunkIndex := LogicChunkIndex(off / cw.ChunkSize)
-	offsetRemainder := off % cw.ChunkSize
+	logicChunkIndex := LogicChunkIndex(off / up.ChunkSize)
 
-	memChunk, found := cw.writableChunks[logicChunkIndex]
+	memChunk, found := up.writableChunks[logicChunkIndex]
 	if !found {
-		memChunk = &MemChunk{
-			buf:   mem.Allocate(int(cw.ChunkSize)),
-			usage: newChunkWrittenIntervalList(),
+		if len(up.writableChunks) < up.bufferChunkLimit {
+			memChunk = NewMemChunk(logicChunkIndex, up.ChunkSize)
+		} else {
+			fullestChunkIndex, fullness := LogicChunkIndex(-1), int64(0)
+			for lci, mc := range up.writableChunks {
+				chunkFullness := mc.WrittenSize()
+				if fullness < chunkFullness {
+					fullestChunkIndex = lci
+					fullness = chunkFullness
+				}
+			}
+			up.moveToSealed(up.writableChunks[fullestChunkIndex], fullestChunkIndex)
+			delete(up.writableChunks, fullestChunkIndex)
+			fmt.Printf("flush chunk %d with %d bytes written", logicChunkIndex, fullness)
+			memChunk = NewMemChunk(logicChunkIndex, up.ChunkSize)
 		}
-		cw.writableChunks[logicChunkIndex] = memChunk
+		up.writableChunks[logicChunkIndex] = memChunk
 	}
-	n = copy(memChunk.buf[offsetRemainder:], p)
-	memChunk.usage.MarkWritten(offsetRemainder, offsetRemainder+int64(n))
-	cw.maybeMoveToSealed(memChunk, logicChunkIndex)
+	n = memChunk.WriteDataAt(p, off)
+	up.maybeMoveToSealed(memChunk, logicChunkIndex)
 
 	return
 }
 
-func (cw *UploadPipeline) MaybeReadDataAt(p []byte, off int64) (maxStop int64) {
-	logicChunkIndex := LogicChunkIndex(off / cw.ChunkSize)
+func (up *UploadPipeline) MaybeReadDataAt(p []byte, off int64) (maxStop int64) {
+	logicChunkIndex := LogicChunkIndex(off / up.ChunkSize)
 
 	// read from sealed chunks first
-	cw.sealedChunksLock.Lock()
-	sealedChunk, found := cw.sealedChunks[logicChunkIndex]
+	up.sealedChunksLock.Lock()
+	sealedChunk, found := up.sealedChunks[logicChunkIndex]
 	if found {
 		sealedChunk.referenceCounter++
 	}
-	cw.sealedChunksLock.Unlock()
+	up.sealedChunksLock.Unlock()
 	if found {
-		maxStop = readMemChunk(sealedChunk.chunk, p, off, logicChunkIndex, cw.ChunkSize)
-		glog.V(4).Infof("%s read sealed memchunk [%d,%d)", cw.filepath, off, maxStop)
-		sealedChunk.FreeReference(fmt.Sprintf("%s finish reading chunk %d", cw.filepath, logicChunkIndex))
+		maxStop = sealedChunk.chunk.ReadDataAt(p, off)
+		glog.V(4).Infof("%s read sealed memchunk [%d,%d)", up.filepath, off, maxStop)
+		sealedChunk.FreeReference(fmt.Sprintf("%s finish reading chunk %d", up.filepath, logicChunkIndex))
 	}
 
 	// read from writable chunks last
-	cw.writableChunksLock.Lock()
-	defer cw.writableChunksLock.Unlock()
-	writableChunk, found := cw.writableChunks[logicChunkIndex]
+	up.writableChunksLock.Lock()
+	defer up.writableChunksLock.Unlock()
+	writableChunk, found := up.writableChunks[logicChunkIndex]
 	if !found {
 		return
 	}
-	writableMaxStop := readMemChunk(writableChunk, p, off, logicChunkIndex, cw.ChunkSize)
-	glog.V(4).Infof("%s read writable memchunk [%d,%d)", cw.filepath, off, writableMaxStop)
+	writableMaxStop := writableChunk.ReadDataAt(p, off)
+	glog.V(4).Infof("%s read writable memchunk [%d,%d)", up.filepath, off, writableMaxStop)
 	maxStop = max(maxStop, writableMaxStop)
 
 	return
 }
 
-func (cw *UploadPipeline) FlushAll() {
-	cw.writableChunksLock.Lock()
-	defer cw.writableChunksLock.Unlock()
+func (up *UploadPipeline) FlushAll() {
+	up.writableChunksLock.Lock()
+	defer up.writableChunksLock.Unlock()
 
-	for logicChunkIndex, memChunk := range cw.writableChunks {
-		cw.moveToSealed(memChunk, logicChunkIndex)
+	for logicChunkIndex, memChunk := range up.writableChunks {
+		up.moveToSealed(memChunk, logicChunkIndex)
 	}
 
-	cw.waitForCurrentWritersToComplete()
+	up.waitForCurrentWritersToComplete()
 }
 
-func (cw *UploadPipeline) LockForRead(startOffset, stopOffset int64) {
-	startLogicChunkIndex := LogicChunkIndex(startOffset / cw.ChunkSize)
-	stopLogicChunkIndex := LogicChunkIndex(stopOffset / cw.ChunkSize)
-	if stopOffset%cw.ChunkSize > 0 {
-		stopLogicChunkIndex += 1
-	}
-	cw.activeReadChunksLock.Lock()
-	defer cw.activeReadChunksLock.Unlock()
-	for i := startLogicChunkIndex; i < stopLogicChunkIndex; i++ {
-		if count, found := cw.activeReadChunks[i]; found {
-			cw.activeReadChunks[i] = count + 1
-		} else {
-			cw.activeReadChunks[i] = 1
-		}
+func (up *UploadPipeline) maybeMoveToSealed(memChunk PageChunk, logicChunkIndex LogicChunkIndex) {
+	if memChunk.IsComplete() {
+		up.moveToSealed(memChunk, logicChunkIndex)
 	}
 }
 
-func (cw *UploadPipeline) UnlockForRead(startOffset, stopOffset int64) {
-	startLogicChunkIndex := LogicChunkIndex(startOffset / cw.ChunkSize)
-	stopLogicChunkIndex := LogicChunkIndex(stopOffset / cw.ChunkSize)
-	if stopOffset%cw.ChunkSize > 0 {
-		stopLogicChunkIndex += 1
-	}
-	cw.activeReadChunksLock.Lock()
-	defer cw.activeReadChunksLock.Unlock()
-	for i := startLogicChunkIndex; i < stopLogicChunkIndex; i++ {
-		if count, found := cw.activeReadChunks[i]; found {
-			if count == 1 {
-				delete(cw.activeReadChunks, i)
-			} else {
-				cw.activeReadChunks[i] = count - 1
-			}
-		}
-	}
-}
+func (up *UploadPipeline) moveToSealed(memChunk PageChunk, logicChunkIndex LogicChunkIndex) {
+	atomic.AddInt32(&up.uploaderCount, 1)
+	glog.V(4).Infof("%s uploaderCount %d ++> %d", up.filepath, up.uploaderCount-1, up.uploaderCount)
 
-func (cw *UploadPipeline) IsLocked(logicChunkIndex LogicChunkIndex) bool {
-	cw.activeReadChunksLock.Lock()
-	defer cw.activeReadChunksLock.Unlock()
-	if count, found := cw.activeReadChunks[logicChunkIndex]; found {
-		return count > 0
-	}
-	return false
-}
+	up.sealedChunksLock.Lock()
 
-func (cw *UploadPipeline) waitForCurrentWritersToComplete() {
-	cw.activeWriterCond.L.Lock()
-	t := int32(100)
-	for {
-		t = atomic.LoadInt32(&cw.activeWriterCount)
-		if t <= 0 {
-			break
-		}
-		glog.V(4).Infof("activeWriterCond is %d", t)
-		cw.activeWriterCond.Wait()
-	}
-	cw.activeWriterCond.L.Unlock()
-}
-
-func (cw *UploadPipeline) maybeMoveToSealed(memChunk *MemChunk, logicChunkIndex LogicChunkIndex) {
-	if memChunk.usage.IsComplete(cw.ChunkSize) {
-		cw.moveToSealed(memChunk, logicChunkIndex)
-	}
-}
-
-func (cw *UploadPipeline) moveToSealed(memChunk *MemChunk, logicChunkIndex LogicChunkIndex) {
-	atomic.AddInt32(&cw.activeWriterCount, 1)
-	glog.V(4).Infof("%s activeWriterCount %d ++> %d", cw.filepath, cw.activeWriterCount-1, cw.activeWriterCount)
-
-	cw.sealedChunksLock.Lock()
-
-	if oldMemChunk, found := cw.sealedChunks[logicChunkIndex]; found {
-		oldMemChunk.FreeReference(fmt.Sprintf("%s replace chunk %d", cw.filepath, logicChunkIndex))
+	if oldMemChunk, found := up.sealedChunks[logicChunkIndex]; found {
+		oldMemChunk.FreeReference(fmt.Sprintf("%s replace chunk %d", up.filepath, logicChunkIndex))
 	}
 	sealedChunk := &SealedChunk{
 		chunk:            memChunk,
 		referenceCounter: 1, // default 1 is for uploading process
 	}
-	cw.sealedChunks[logicChunkIndex] = sealedChunk
-	delete(cw.writableChunks, logicChunkIndex)
+	up.sealedChunks[logicChunkIndex] = sealedChunk
+	delete(up.writableChunks, logicChunkIndex)
 
-	cw.sealedChunksLock.Unlock()
+	up.sealedChunksLock.Unlock()
 
-	cw.writers.Execute(func() {
+	up.uploaders.Execute(func() {
 		// first add to the file chunks
-		cw.saveOneChunk(sealedChunk.chunk, logicChunkIndex)
+		sealedChunk.chunk.SaveContent(up.saveToStorageFn)
 
 		// notify waiting process
-		atomic.AddInt32(&cw.activeWriterCount, -1)
-		glog.V(4).Infof("%s activeWriterCount %d --> %d", cw.filepath, cw.activeWriterCount+1, cw.activeWriterCount)
+		atomic.AddInt32(&up.uploaderCount, -1)
+		glog.V(4).Infof("%s uploaderCount %d --> %d", up.filepath, up.uploaderCount+1, up.uploaderCount)
 		// Lock and Unlock are not required,
 		// but it may signal multiple times during one wakeup,
 		// and the waiting goroutine may miss some of them!
-		cw.activeWriterCond.L.Lock()
-		cw.activeWriterCond.Broadcast()
-		cw.activeWriterCond.L.Unlock()
+		up.uploaderCountCond.L.Lock()
+		up.uploaderCountCond.Broadcast()
+		up.uploaderCountCond.L.Unlock()
 
 		// wait for readers
-		for cw.IsLocked(logicChunkIndex) {
+		for up.IsLocked(logicChunkIndex) {
 			time.Sleep(59 * time.Millisecond)
 		}
 
 		// then remove from sealed chunks
-		cw.sealedChunksLock.Lock()
-		defer cw.sealedChunksLock.Unlock()
-		delete(cw.sealedChunks, logicChunkIndex)
-		sealedChunk.FreeReference(fmt.Sprintf("%s finished uploading chunk %d", cw.filepath, logicChunkIndex))
+		up.sealedChunksLock.Lock()
+		defer up.sealedChunksLock.Unlock()
+		delete(up.sealedChunks, logicChunkIndex)
+		sealedChunk.FreeReference(fmt.Sprintf("%s finished uploading chunk %d", up.filepath, logicChunkIndex))
 
 	})
 }
 
-func (cw *UploadPipeline) saveOneChunk(memChunk *MemChunk, logicChunkIndex LogicChunkIndex) {
-	if cw.saveToStorageFn == nil {
-		return
-	}
-	for t := memChunk.usage.head.next; t != memChunk.usage.tail; t = t.next {
-		reader := util.NewBytesReader(memChunk.buf[t.StartOffset:t.stopOffset])
-		cw.saveToStorageFn(reader, int64(logicChunkIndex)*cw.ChunkSize+t.StartOffset, t.Size(), func() {
-		})
-	}
-}
-
-func readMemChunk(memChunk *MemChunk, p []byte, off int64, logicChunkIndex LogicChunkIndex, chunkSize int64) (maxStop int64) {
-	memChunkBaseOffset := int64(logicChunkIndex) * chunkSize
-	for t := memChunk.usage.head.next; t != memChunk.usage.tail; t = t.next {
-		logicStart := max(off, int64(logicChunkIndex)*chunkSize+t.StartOffset)
-		logicStop := min(off+int64(len(p)), memChunkBaseOffset+t.stopOffset)
-		if logicStart < logicStop {
-			copy(p[logicStart-off:logicStop-off], memChunk.buf[logicStart-memChunkBaseOffset:logicStop-memChunkBaseOffset])
-			maxStop = max(maxStop, logicStop)
-		}
-	}
-	return
-}
-
-func (p2 *UploadPipeline) Shutdown() {
-
+func (up *UploadPipeline) Shutdown() {
 }

weed/filesys/page_writer/upload_pipeline_lock.go

@@ -0,0 +1,63 @@
+package page_writer
+
+import (
+	"sync/atomic"
+)
+
+func (up *UploadPipeline) LockForRead(startOffset, stopOffset int64) {
+	startLogicChunkIndex := LogicChunkIndex(startOffset / up.ChunkSize)
+	stopLogicChunkIndex := LogicChunkIndex(stopOffset / up.ChunkSize)
+	if stopOffset%up.ChunkSize > 0 {
+		stopLogicChunkIndex += 1
+	}
+	up.activeReadChunksLock.Lock()
+	defer up.activeReadChunksLock.Unlock()
+	for i := startLogicChunkIndex; i < stopLogicChunkIndex; i++ {
+		if count, found := up.activeReadChunks[i]; found {
+			up.activeReadChunks[i] = count + 1
+		} else {
+			up.activeReadChunks[i] = 1
+		}
+	}
+}
+
+func (up *UploadPipeline) UnlockForRead(startOffset, stopOffset int64) {
+	startLogicChunkIndex := LogicChunkIndex(startOffset / up.ChunkSize)
+	stopLogicChunkIndex := LogicChunkIndex(stopOffset / up.ChunkSize)
+	if stopOffset%up.ChunkSize > 0 {
+		stopLogicChunkIndex += 1
+	}
+	up.activeReadChunksLock.Lock()
+	defer up.activeReadChunksLock.Unlock()
+	for i := startLogicChunkIndex; i < stopLogicChunkIndex; i++ {
+		if count, found := up.activeReadChunks[i]; found {
+			if count == 1 {
+				delete(up.activeReadChunks, i)
+			} else {
+				up.activeReadChunks[i] = count - 1
+			}
+		}
+	}
+}
+
+func (up *UploadPipeline) IsLocked(logicChunkIndex LogicChunkIndex) bool {
+	up.activeReadChunksLock.Lock()
+	defer up.activeReadChunksLock.Unlock()
+	if count, found := up.activeReadChunks[logicChunkIndex]; found {
+		return count > 0
+	}
+	return false
+}
+
+func (up *UploadPipeline) waitForCurrentWritersToComplete() {
+	up.uploaderCountCond.L.Lock()
+	t := int32(100)
+	for {
+		t = atomic.LoadInt32(&up.uploaderCount)
+		if t <= 0 {
+			break
+		}
+		up.uploaderCountCond.Wait()
+	}
+	up.uploaderCountCond.L.Unlock()
+}

weed/filesys/page_writer/upload_pipeline_test.go

@@ -7,7 +7,7 @@ import (
 
 func TestUploadPipeline(t *testing.T) {
 
-	uploadPipeline := NewUploadPipeline("", nil, 2*1024*1024, nil)
+	uploadPipeline := NewUploadPipeline(nil, 2*1024*1024, nil, 16)
 
 	writeRange(uploadPipeline, 0, 131072)
 	writeRange(uploadPipeline, 131072, 262144)

weed/images/orientation_test.go

@@ -1,6 +1,7 @@
 package images
 
 import (
+	"github.com/chrislusf/seaweedfs/weed/util"
 	"os"
 	"testing"
 )
@@ -12,7 +13,7 @@ func TestXYZ(t *testing.T) {
 
 	fixed_data := FixJpgOrientation(dat)
 
-	os.WriteFile("fixed1.jpg", fixed_data, 0644)
+	util.WriteFile("fixed1.jpg", fixed_data, 0644)
 
 	os.Remove("fixed1.jpg")
 

weed/images/resizing_test.go

@@ -2,6 +2,7 @@ package images
 
 import (
 	"bytes"
+	"github.com/chrislusf/seaweedfs/weed/util"
 	"os"
 	"testing"
 )
@@ -15,7 +16,7 @@ func TestResizing(t *testing.T) {
 	buf := new(bytes.Buffer)
 	buf.ReadFrom(resized)
 
-	os.WriteFile("resized1.png", buf.Bytes(), 0644)
+	util.WriteFile("resized1.png", buf.Bytes(), 0644)
 
 	os.Remove("resized1.png")
 

weed/mount/directory.go

@@ -0,0 +1,42 @@
+package mount
+
+import (
+	"bytes"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/hanwen/go-fuse/v2/fs"
+	"strings"
+)
+
+type Directory struct {
+	fs.Inode
+
+	name   string
+	wfs    *WFS
+	entry  *filer_pb.Entry
+	parent *Directory
+	id     uint64
+}
+
+func (dir *Directory) FullPath() string {
+	var parts []string
+	for p := dir; p != nil; p = p.parent {
+		if strings.HasPrefix(p.name, "/") {
+			if len(p.name) > 1 {
+				parts = append(parts, p.name[1:])
+			}
+		} else {
+			parts = append(parts, p.name)
+		}
+	}
+
+	if len(parts) == 0 {
+		return "/"
+	}
+
+	var buf bytes.Buffer
+	for i := len(parts) - 1; i >= 0; i-- {
+		buf.WriteString("/")
+		buf.WriteString(parts[i])
+	}
+	return buf.String()
+}

weed/mount/directory_read.go

@@ -0,0 +1,84 @@
+package mount
+
+import (
+	"context"
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/mount/meta_cache"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"github.com/hanwen/go-fuse/v2/fs"
+	"github.com/hanwen/go-fuse/v2/fuse"
+	"math"
+	"os"
+	"syscall"
+)
+
+var _ = fs.NodeReaddirer(&Directory{})
+var _ = fs.NodeGetattrer(&Directory{})
+
+func (dir *Directory) Getattr(ctx context.Context, fh fs.FileHandle, out *fuse.AttrOut) syscall.Errno {
+	out.Mode = 0755
+	return 0
+}
+
+func (dir *Directory) Readdir(ctx context.Context) (fs.DirStream, syscall.Errno) {
+
+	dirPath := util.FullPath(dir.FullPath())
+	glog.V(4).Infof("Readdir %s", dirPath)
+
+	sourceChan := make(chan fuse.DirEntry, 64)
+
+	stream := newDirectoryListStream(sourceChan)
+
+	processEachEntryFn := func(entry *filer.Entry, isLast bool) {
+		sourceChan <- fuse.DirEntry{
+			Mode: uint32(entry.Mode),
+			Name: entry.Name(),
+			Ino:  dirPath.Child(entry.Name()).AsInode(os.ModeDir),
+		}
+	}
+
+	if err := meta_cache.EnsureVisited(dir.wfs.metaCache, dir.wfs, dirPath); err != nil {
+		glog.Errorf("dir ReadDirAll %s: %v", dirPath, err)
+		return nil, fs.ToErrno(os.ErrInvalid)
+	}
+	go func() {
+		dir.wfs.metaCache.ListDirectoryEntries(context.Background(), dirPath, "", false, int64(math.MaxInt32), func(entry *filer.Entry) bool {
+			processEachEntryFn(entry, false)
+			return true
+		})
+		close(sourceChan)
+	}()
+
+	return stream, fs.OK
+}
+
+var _ = fs.DirStream(&DirectoryListStream{})
+
+type DirectoryListStream struct {
+	next       fuse.DirEntry
+	sourceChan chan fuse.DirEntry
+	isStarted  bool
+	hasNext    bool
+}
+
+func newDirectoryListStream(ch chan fuse.DirEntry) *DirectoryListStream {
+	return &DirectoryListStream{
+		sourceChan: ch,
+	}
+}
+
+func (i *DirectoryListStream) HasNext() bool {
+	if !i.isStarted {
+		i.next, i.hasNext = <-i.sourceChan
+		i.isStarted = true
+	}
+	return i.hasNext
+}
+func (i *DirectoryListStream) Next() (fuse.DirEntry, syscall.Errno) {
+	t := i.next
+	i.next, i.hasNext = <-i.sourceChan
+	return t, fs.OK
+}
+func (i *DirectoryListStream) Close() {
+}

weed/mount/dirty_pages_chunked.go

@@ -0,0 +1,99 @@
+package mount
+
+import (
+	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/filesys/page_writer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"io"
+	"sync"
+	"time"
+)
+
+type ChunkedDirtyPages struct {
+	fh             *FileHandle
+	writeWaitGroup sync.WaitGroup
+	lastErr        error
+	collection     string
+	replication    string
+	uploadPipeline *page_writer.UploadPipeline
+	hasWrites      bool
+}
+
+var (
+	_ = page_writer.DirtyPages(&ChunkedDirtyPages{})
+)
+
+func newMemoryChunkPages(fh *FileHandle, chunkSize int64) *ChunkedDirtyPages {
+
+	dirtyPages := &ChunkedDirtyPages{
+		fh: fh,
+	}
+
+	dirtyPages.uploadPipeline = page_writer.NewUploadPipeline(fh.wfs.concurrentWriters, chunkSize, dirtyPages.saveChunkedFileIntevalToStorage, fh.wfs.option.ConcurrentWriters)
+
+	return dirtyPages
+}
+
+func (pages *ChunkedDirtyPages) AddPage(offset int64, data []byte) {
+	pages.hasWrites = true
+
+	glog.V(4).Infof("%v memory AddPage [%d, %d)", pages.fh, offset, offset+int64(len(data)))
+	pages.uploadPipeline.SaveDataAt(data, offset)
+
+	return
+}
+
+func (pages *ChunkedDirtyPages) FlushData() error {
+	if !pages.hasWrites {
+		return nil
+	}
+	pages.uploadPipeline.FlushAll()
+	if pages.lastErr != nil {
+		return fmt.Errorf("flush data: %v", pages.lastErr)
+	}
+	return nil
+}
+
+func (pages *ChunkedDirtyPages) ReadDirtyDataAt(data []byte, startOffset int64) (maxStop int64) {
+	if !pages.hasWrites {
+		return
+	}
+	return pages.uploadPipeline.MaybeReadDataAt(data, startOffset)
+}
+
+func (pages *ChunkedDirtyPages) GetStorageOptions() (collection, replication string) {
+	return pages.collection, pages.replication
+}
+
+func (pages *ChunkedDirtyPages) saveChunkedFileIntevalToStorage(reader io.Reader, offset int64, size int64, cleanupFn func()) {
+
+	mtime := time.Now().UnixNano()
+	defer cleanupFn()
+
+	fileFullPath := pages.fh.FullPath()
+	fileName := fileFullPath.Name()
+	chunk, collection, replication, err := pages.fh.wfs.saveDataAsChunk(fileFullPath)(reader, fileName, offset)
+	if err != nil {
+		glog.V(0).Infof("%v saveToStorage [%d,%d): %v", fileFullPath, offset, offset+size, err)
+		pages.lastErr = err
+		return
+	}
+	chunk.Mtime = mtime
+	pages.collection, pages.replication = collection, replication
+	pages.fh.addChunks([]*filer_pb.FileChunk{chunk})
+	pages.fh.entryViewCache = nil
+	glog.V(3).Infof("%v saveToStorage %s [%d,%d)", fileFullPath, chunk.FileId, offset, offset+size)
+
+}
+
+func (pages ChunkedDirtyPages) Destroy() {
+	pages.uploadPipeline.Shutdown()
+}
+
+func (pages *ChunkedDirtyPages) LockForRead(startOffset, stopOffset int64) {
+	pages.uploadPipeline.LockForRead(startOffset, stopOffset)
+}
+func (pages *ChunkedDirtyPages) UnlockForRead(startOffset, stopOffset int64) {
+	pages.uploadPipeline.UnlockForRead(startOffset, stopOffset)
+}

weed/mount/filehandle.go

@@ -0,0 +1,95 @@
+package mount
+
+import (
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"io"
+	"sort"
+	"sync"
+)
+
+type FileHandleId uint64
+
+type FileHandle struct {
+	fh           FileHandleId
+	counter      int64
+	entry        *filer_pb.Entry
+	chunkAddLock sync.Mutex
+	inode        uint64
+	wfs          *WFS
+
+	// cache file has been written to
+	dirtyMetadata  bool
+	dirtyPages     *PageWriter
+	entryViewCache []filer.VisibleInterval
+	reader         io.ReaderAt
+	contentType    string
+	handle         uint64
+	sync.Mutex
+
+	isDeleted bool
+}
+
+func newFileHandle(wfs *WFS, handleId FileHandleId, inode uint64, entry *filer_pb.Entry) *FileHandle {
+	fh := &FileHandle{
+		fh:      handleId,
+		counter: 1,
+		inode:   inode,
+		wfs:     wfs,
+	}
+	// dirtyPages: newContinuousDirtyPages(file, writeOnly),
+	fh.dirtyPages = newPageWriter(fh, wfs.option.ChunkSizeLimit)
+	if entry != nil {
+		entry.Attributes.FileSize = filer.FileSize(entry)
+	}
+
+	return fh
+}
+
+func (fh *FileHandle) FullPath() util.FullPath {
+	return fh.wfs.inodeToPath.GetPath(fh.inode)
+}
+
+func (fh *FileHandle) addChunks(chunks []*filer_pb.FileChunk) {
+
+	// find the earliest incoming chunk
+	newChunks := chunks
+	earliestChunk := newChunks[0]
+	for i := 1; i < len(newChunks); i++ {
+		if lessThan(earliestChunk, newChunks[i]) {
+			earliestChunk = newChunks[i]
+		}
+	}
+
+	if fh.entry == nil {
+		return
+	}
+
+	// pick out-of-order chunks from existing chunks
+	for _, chunk := range fh.entry.Chunks {
+		if lessThan(earliestChunk, chunk) {
+			chunks = append(chunks, chunk)
+		}
+	}
+
+	// sort incoming chunks
+	sort.Slice(chunks, func(i, j int) bool {
+		return lessThan(chunks[i], chunks[j])
+	})
+
+	glog.V(4).Infof("%s existing %d chunks adds %d more", fh.FullPath(), len(fh.entry.Chunks), len(chunks))
+
+	fh.chunkAddLock.Lock()
+	fh.entry.Chunks = append(fh.entry.Chunks, newChunks...)
+	fh.entryViewCache = nil
+	fh.chunkAddLock.Unlock()
+}
+
+func lessThan(a, b *filer_pb.FileChunk) bool {
+	if a.Mtime == b.Mtime {
+		return a.Fid.FileKey < b.Fid.FileKey
+	}
+	return a.Mtime < b.Mtime
+}

weed/mount/filehandle_map.go

@@ -0,0 +1,84 @@
+package mount
+
+import (
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"sync"
+)
+
+type FileHandleToInode struct {
+	sync.RWMutex
+	nextFh   FileHandleId
+	inode2fh map[uint64]*FileHandle
+	fh2inode map[FileHandleId]uint64
+}
+
+func NewFileHandleToInode() *FileHandleToInode {
+	return &FileHandleToInode{
+		inode2fh: make(map[uint64]*FileHandle),
+		fh2inode: make(map[FileHandleId]uint64),
+		nextFh:   0,
+	}
+}
+
+func (i *FileHandleToInode) GetFileHandle(fh FileHandleId) *FileHandle {
+	i.RLock()
+	defer i.RUnlock()
+	inode, found := i.fh2inode[fh]
+	if found {
+		return i.inode2fh[inode]
+	}
+	return nil
+}
+
+func (i *FileHandleToInode) FindFileHandle(inode uint64) (fh *FileHandle, found bool) {
+	i.RLock()
+	defer i.RUnlock()
+	fh, found = i.inode2fh[inode]
+	return
+}
+
+func (i *FileHandleToInode) AcquireFileHandle(wfs *WFS, inode uint64, entry *filer_pb.Entry) *FileHandle {
+	i.Lock()
+	defer i.Unlock()
+	fh, found := i.inode2fh[inode]
+	if !found {
+		fh = newFileHandle(wfs, i.nextFh, inode, entry)
+		i.nextFh++
+		i.inode2fh[inode] = fh
+		i.fh2inode[fh.fh] = inode
+	} else {
+		fh.counter++
+	}
+	return fh
+}
+
+func (i *FileHandleToInode) ReleaseByInode(inode uint64) {
+	i.Lock()
+	defer i.Unlock()
+	fh, found := i.inode2fh[inode]
+	if found {
+		fh.counter--
+		if fh.counter <= 0 {
+			delete(i.inode2fh, inode)
+			delete(i.fh2inode, fh.fh)
+		}
+	}
+}
+func (i *FileHandleToInode) ReleaseByHandle(fh FileHandleId) {
+	i.Lock()
+	defer i.Unlock()
+	inode, found := i.fh2inode[fh]
+	if found {
+		fhHandle, fhFound := i.inode2fh[inode]
+		if !fhFound {
+			delete(i.fh2inode, fh)
+		} else {
+			fhHandle.counter--
+			if fhHandle.counter <= 0 {
+				delete(i.inode2fh, inode)
+				delete(i.fh2inode, fhHandle.fh)
+			}
+		}
+
+	}
+}

weed/mount/filehandle_read.go

@@ -0,0 +1,114 @@
+package mount
+
+import (
+	"context"
+	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"io"
+	"math"
+)
+
+func (fh *FileHandle) lockForRead(startOffset int64, size int) {
+	fh.dirtyPages.LockForRead(startOffset, startOffset+int64(size))
+}
+func (fh *FileHandle) unlockForRead(startOffset int64, size int) {
+	fh.dirtyPages.UnlockForRead(startOffset, startOffset+int64(size))
+}
+
+func (fh *FileHandle) readFromDirtyPages(buff []byte, startOffset int64) (maxStop int64) {
+	maxStop = fh.dirtyPages.ReadDirtyDataAt(buff, startOffset)
+	return
+}
+
+func (fh *FileHandle) readFromChunks(buff []byte, offset int64) (int64, error) {
+
+	fileFullPath := fh.FullPath()
+
+	entry := fh.entry
+	if entry == nil {
+		return 0, io.EOF
+	}
+
+	if entry.IsInRemoteOnly() {
+		glog.V(4).Infof("download remote entry %s", fileFullPath)
+		newEntry, err := fh.downloadRemoteEntry(entry)
+		if err != nil {
+			glog.V(1).Infof("download remote entry %s: %v", fileFullPath, err)
+			return 0, err
+		}
+		entry = newEntry
+	}
+
+	fileSize := int64(filer.FileSize(entry))
+
+	if fileSize == 0 {
+		glog.V(1).Infof("empty fh %v", fileFullPath)
+		return 0, io.EOF
+	}
+
+	if offset+int64(len(buff)) <= int64(len(entry.Content)) {
+		totalRead := copy(buff, entry.Content[offset:])
+		glog.V(4).Infof("file handle read cached %s [%d,%d] %d", fileFullPath, offset, offset+int64(totalRead), totalRead)
+		return int64(totalRead), nil
+	}
+
+	var chunkResolveErr error
+	if fh.entryViewCache == nil {
+		fh.entryViewCache, chunkResolveErr = filer.NonOverlappingVisibleIntervals(fh.wfs.LookupFn(), entry.Chunks, 0, math.MaxInt64)
+		if chunkResolveErr != nil {
+			return 0, fmt.Errorf("fail to resolve chunk manifest: %v", chunkResolveErr)
+		}
+		fh.reader = nil
+	}
+
+	reader := fh.reader
+	if reader == nil {
+		chunkViews := filer.ViewFromVisibleIntervals(fh.entryViewCache, 0, math.MaxInt64)
+		glog.V(4).Infof("file handle read %s [%d,%d) from %d views", fileFullPath, offset, offset+int64(len(buff)), len(chunkViews))
+		for _, chunkView := range chunkViews {
+			glog.V(4).Infof("  read %s [%d,%d) from chunk %+v", fileFullPath, chunkView.LogicOffset, chunkView.LogicOffset+int64(chunkView.Size), chunkView.FileId)
+		}
+		reader = filer.NewChunkReaderAtFromClient(fh.wfs.LookupFn(), chunkViews, fh.wfs.chunkCache, fileSize)
+	}
+	fh.reader = reader
+
+	totalRead, err := reader.ReadAt(buff, offset)
+
+	if err != nil && err != io.EOF {
+		glog.Errorf("file handle read %s: %v", fileFullPath, err)
+	}
+
+	glog.V(4).Infof("file handle read %s [%d,%d] %d : %v", fileFullPath, offset, offset+int64(totalRead), totalRead, err)
+
+	return int64(totalRead), err
+}
+
+func (fh *FileHandle) downloadRemoteEntry(entry *filer_pb.Entry) (*filer_pb.Entry, error) {
+
+	fileFullPath := fh.FullPath()
+	dir, _ := fileFullPath.DirAndName()
+
+	err := fh.wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
+
+		request := &filer_pb.CacheRemoteObjectToLocalClusterRequest{
+			Directory: string(dir),
+			Name:      entry.Name,
+		}
+
+		glog.V(4).Infof("download entry: %v", request)
+		resp, err := client.CacheRemoteObjectToLocalCluster(context.Background(), request)
+		if err != nil {
+			return fmt.Errorf("CacheRemoteObjectToLocalCluster file %s: %v", fileFullPath, err)
+		}
+
+		entry = resp.Entry
+
+		fh.wfs.metaCache.InsertEntry(context.Background(), filer.FromPbEntry(request.Directory, resp.Entry))
+
+		return nil
+	})
+
+	return entry, err
+}

weed/mount/inode_to_path.go

@@ -0,0 +1,161 @@
+package mount
+
+import (
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"sync"
+)
+
+type InodeToPath struct {
+	sync.RWMutex
+	nextInodeId uint64
+	inode2path  map[uint64]*InodeEntry
+	path2inode  map[util.FullPath]uint64
+}
+type InodeEntry struct {
+	util.FullPath
+	nlookup          uint64
+	isDirectory      bool
+	isChildrenCached bool
+}
+
+func NewInodeToPath() *InodeToPath {
+	t := &InodeToPath{
+		inode2path:  make(map[uint64]*InodeEntry),
+		path2inode:  make(map[util.FullPath]uint64),
+		nextInodeId: 2, // the root inode id is 1
+	}
+	t.inode2path[1] = &InodeEntry{"/", 1, true, false}
+	t.path2inode["/"] = 1
+	return t
+}
+
+func (i *InodeToPath) Lookup(path util.FullPath, isDirectory bool) uint64 {
+	i.Lock()
+	defer i.Unlock()
+	inode, found := i.path2inode[path]
+	if !found {
+		inode = i.nextInodeId
+		i.nextInodeId++
+		i.path2inode[path] = inode
+		i.inode2path[inode] = &InodeEntry{path, 1, isDirectory, false}
+	} else {
+		i.inode2path[inode].nlookup++
+	}
+	return inode
+}
+
+func (i *InodeToPath) GetInode(path util.FullPath) uint64 {
+	if path == "/" {
+		return 1
+	}
+	i.Lock()
+	defer i.Unlock()
+	inode, found := i.path2inode[path]
+	if !found {
+		// glog.Fatalf("GetInode unknown inode for %s", path)
+		// this could be the parent for mount point
+	}
+	return inode
+}
+
+func (i *InodeToPath) GetPath(inode uint64) util.FullPath {
+	i.RLock()
+	defer i.RUnlock()
+	path, found := i.inode2path[inode]
+	if !found {
+		glog.Fatalf("not found inode %d", inode)
+	}
+	return path.FullPath
+}
+
+func (i *InodeToPath) HasPath(path util.FullPath) bool {
+	i.RLock()
+	defer i.RUnlock()
+	_, found := i.path2inode[path]
+	return found
+}
+
+func (i *InodeToPath) MarkChildrenCached(fullpath util.FullPath) {
+	i.RLock()
+	defer i.RUnlock()
+	inode, found := i.path2inode[fullpath]
+	if !found {
+		glog.Fatalf("MarkChildrenCached not found inode %v", fullpath)
+	}
+	path, found := i.inode2path[inode]
+	path.isChildrenCached = true
+}
+
+func (i *InodeToPath) IsChildrenCached(fullpath util.FullPath) bool {
+	i.RLock()
+	defer i.RUnlock()
+	inode, found := i.path2inode[fullpath]
+	if !found {
+		return false
+	}
+	path, found := i.inode2path[inode]
+	if found {
+		return path.isChildrenCached
+	}
+	return false
+}
+
+func (i *InodeToPath) HasInode(inode uint64) bool {
+	if inode == 1 {
+		return true
+	}
+	i.RLock()
+	defer i.RUnlock()
+	_, found := i.inode2path[inode]
+	return found
+}
+
+func (i *InodeToPath) RemovePath(path util.FullPath) {
+	i.Lock()
+	defer i.Unlock()
+	inode, found := i.path2inode[path]
+	if found {
+		delete(i.path2inode, path)
+		delete(i.inode2path, inode)
+	}
+}
+
+func (i *InodeToPath) MovePath(sourcePath, targetPath util.FullPath) {
+	i.Lock()
+	defer i.Unlock()
+	sourceInode, sourceFound := i.path2inode[sourcePath]
+	targetInode, targetFound := i.path2inode[targetPath]
+	if sourceFound {
+		delete(i.path2inode, sourcePath)
+		i.path2inode[targetPath] = sourceInode
+	} else {
+		// it is possible some source folder items has not been visited before
+		// so no need to worry about their source inodes
+		return
+	}
+	i.inode2path[sourceInode].FullPath = targetPath
+	if targetFound {
+		delete(i.inode2path, targetInode)
+	} else {
+		i.inode2path[sourceInode].nlookup++
+	}
+}
+
+func (i *InodeToPath) Forget(inode, nlookup uint64, onForgetDir func(dir util.FullPath)) {
+	i.Lock()
+	path, found := i.inode2path[inode]
+	if found {
+		path.nlookup -= nlookup
+		if path.nlookup <= 0 {
+			delete(i.path2inode, path.FullPath)
+			delete(i.inode2path, inode)
+		}
+	}
+	i.Unlock()
+	if found {
+		if path.isDirectory && onForgetDir != nil {
+			onForgetDir(path.FullPath)
+		}
+	}
+}

weed/mount/meta_cache/cache_config.go

@@ -0,0 +1,32 @@
+package meta_cache
+
+import "github.com/chrislusf/seaweedfs/weed/util"
+
+var (
+	_ = util.Configuration(&cacheConfig{})
+)
+
+// implementing util.Configuraion
+type cacheConfig struct {
+	dir string
+}
+
+func (c cacheConfig) GetString(key string) string {
+	return c.dir
+}
+
+func (c cacheConfig) GetBool(key string) bool {
+	panic("implement me")
+}
+
+func (c cacheConfig) GetInt(key string) int {
+	panic("implement me")
+}
+
+func (c cacheConfig) GetStringSlice(key string) []string {
+	panic("implement me")
+}
+
+func (c cacheConfig) SetDefault(key string, value interface{}) {
+	panic("implement me")
+}

weed/mount/meta_cache/id_mapper.go

@@ -0,0 +1,101 @@
+package meta_cache
+
+import (
+	"fmt"
+	"strconv"
+	"strings"
+)
+
+type UidGidMapper struct {
+	uidMapper *IdMapper
+	gidMapper *IdMapper
+}
+
+type IdMapper struct {
+	localToFiler map[uint32]uint32
+	filerToLocal map[uint32]uint32
+}
+
+// UidGidMapper translates local uid/gid to filer uid/gid
+// The local storage always persists the same as the filer.
+// The local->filer translation happens when updating the filer first and later saving to meta_cache.
+// And filer->local happens when reading from the meta_cache.
+func NewUidGidMapper(uidPairsStr, gidPairStr string) (*UidGidMapper, error) {
+	uidMapper, err := newIdMapper(uidPairsStr)
+	if err != nil {
+		return nil, err
+	}
+	gidMapper, err := newIdMapper(gidPairStr)
+	if err != nil {
+		return nil, err
+	}
+
+	return &UidGidMapper{
+		uidMapper: uidMapper,
+		gidMapper: gidMapper,
+	}, nil
+}
+
+func (m *UidGidMapper) LocalToFiler(uid, gid uint32) (uint32, uint32) {
+	return m.uidMapper.LocalToFiler(uid), m.gidMapper.LocalToFiler(gid)
+}
+func (m *UidGidMapper) FilerToLocal(uid, gid uint32) (uint32, uint32) {
+	return m.uidMapper.FilerToLocal(uid), m.gidMapper.FilerToLocal(gid)
+}
+
+func (m *IdMapper) LocalToFiler(id uint32) uint32 {
+	value, found := m.localToFiler[id]
+	if found {
+		return value
+	}
+	return id
+}
+func (m *IdMapper) FilerToLocal(id uint32) uint32 {
+	value, found := m.filerToLocal[id]
+	if found {
+		return value
+	}
+	return id
+}
+
+func newIdMapper(pairsStr string) (*IdMapper, error) {
+
+	localToFiler, filerToLocal, err := parseUint32Pairs(pairsStr)
+	if err != nil {
+		return nil, err
+	}
+
+	return &IdMapper{
+		localToFiler: localToFiler,
+		filerToLocal: filerToLocal,
+	}, nil
+
+}
+
+func parseUint32Pairs(pairsStr string) (localToFiler, filerToLocal map[uint32]uint32, err error) {
+
+	if pairsStr == "" {
+		return
+	}
+
+	localToFiler = make(map[uint32]uint32)
+	filerToLocal = make(map[uint32]uint32)
+	for _, pairStr := range strings.Split(pairsStr, ",") {
+		pair := strings.Split(pairStr, ":")
+		localUidStr, filerUidStr := pair[0], pair[1]
+		localUid, localUidErr := strconv.Atoi(localUidStr)
+		if localUidErr != nil {
+			err = fmt.Errorf("failed to parse local %s: %v", localUidStr, localUidErr)
+			return
+		}
+		filerUid, filerUidErr := strconv.Atoi(filerUidStr)
+		if filerUidErr != nil {
+			err = fmt.Errorf("failed to parse remote %s: %v", filerUidStr, filerUidErr)
+			return
+		}
+		localToFiler[uint32(localUid)] = uint32(filerUid)
+		filerToLocal[uint32(filerUid)] = uint32(localUid)
+	}
+
+	return
+}

weed/mount/meta_cache/meta_cache.go

@@ -0,0 +1,160 @@
+package meta_cache
+
+import (
+	"context"
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/filer/leveldb"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"os"
+)
+
+// need to have logic similar to FilerStoreWrapper
+// e.g. fill fileId field for chunks
+
+type MetaCache struct {
+	localStore filer.VirtualFilerStore
+	// sync.RWMutex
+	uidGidMapper   *UidGidMapper
+	markCachedFn   func(fullpath util.FullPath)
+	isCachedFn     func(fullpath util.FullPath) bool
+	invalidateFunc func(fullpath util.FullPath, entry *filer_pb.Entry)
+}
+
+func NewMetaCache(dbFolder string, uidGidMapper *UidGidMapper, markCachedFn func(path util.FullPath), isCachedFn func(path util.FullPath) bool, invalidateFunc func(util.FullPath, *filer_pb.Entry)) *MetaCache {
+	return &MetaCache{
+		localStore:   openMetaStore(dbFolder),
+		markCachedFn: markCachedFn,
+		isCachedFn:   isCachedFn,
+		uidGidMapper: uidGidMapper,
+		invalidateFunc: func(fullpath util.FullPath, entry *filer_pb.Entry) {
+			invalidateFunc(fullpath, entry)
+		},
+	}
+}
+
+func openMetaStore(dbFolder string) filer.VirtualFilerStore {
+
+	os.RemoveAll(dbFolder)
+	os.MkdirAll(dbFolder, 0755)
+
+	store := &leveldb.LevelDBStore{}
+	config := &cacheConfig{
+		dir: dbFolder,
+	}
+
+	if err := store.Initialize(config, ""); err != nil {
+		glog.Fatalf("Failed to initialize metadata cache store for %s: %+v", store.GetName(), err)
+	}
+
+	return filer.NewFilerStoreWrapper(store)
+
+}
+
+func (mc *MetaCache) InsertEntry(ctx context.Context, entry *filer.Entry) error {
+	//mc.Lock()
+	//defer mc.Unlock()
+	return mc.doInsertEntry(ctx, entry)
+}
+
+func (mc *MetaCache) doInsertEntry(ctx context.Context, entry *filer.Entry) error {
+	return mc.localStore.InsertEntry(ctx, entry)
+}
+
+func (mc *MetaCache) AtomicUpdateEntryFromFiler(ctx context.Context, oldPath util.FullPath, newEntry *filer.Entry) error {
+	//mc.Lock()
+	//defer mc.Unlock()
+
+	oldDir, _ := oldPath.DirAndName()
+	if mc.isCachedFn(util.FullPath(oldDir)) {
+		if oldPath != "" {
+			if newEntry != nil && oldPath == newEntry.FullPath {
+				// skip the unnecessary deletion
+				// leave the update to the following InsertEntry operation
+			} else {
+				glog.V(3).Infof("DeleteEntry %s", oldPath)
+				if err := mc.localStore.DeleteEntry(ctx, oldPath); err != nil {
+					return err
+				}
+			}
+		}
+	} else {
+		// println("unknown old directory:", oldDir)
+	}
+
+	if newEntry != nil {
+		newDir, _ := newEntry.DirAndName()
+		if mc.isCachedFn(util.FullPath(newDir)) {
+			glog.V(3).Infof("InsertEntry %s/%s", newDir, newEntry.Name())
+			if err := mc.localStore.InsertEntry(ctx, newEntry); err != nil {
+				return err
+			}
+		}
+	}
+	return nil
+}
+
+func (mc *MetaCache) UpdateEntry(ctx context.Context, entry *filer.Entry) error {
+	//mc.Lock()
+	//defer mc.Unlock()
+	return mc.localStore.UpdateEntry(ctx, entry)
+}
+
+func (mc *MetaCache) FindEntry(ctx context.Context, fp util.FullPath) (entry *filer.Entry, err error) {
+	//mc.RLock()
+	//defer mc.RUnlock()
+	entry, err = mc.localStore.FindEntry(ctx, fp)
+	if err != nil {
+		return nil, err
+	}
+	mc.mapIdFromFilerToLocal(entry)
+	return
+}
+
+func (mc *MetaCache) DeleteEntry(ctx context.Context, fp util.FullPath) (err error) {
+	//mc.Lock()
+	//defer mc.Unlock()
+	return mc.localStore.DeleteEntry(ctx, fp)
+}
+func (mc *MetaCache) DeleteFolderChildren(ctx context.Context, fp util.FullPath) (err error) {
+	//mc.Lock()
+	//defer mc.Unlock()
+	return mc.localStore.DeleteFolderChildren(ctx, fp)
+}
+
+func (mc *MetaCache) ListDirectoryEntries(ctx context.Context, dirPath util.FullPath, startFileName string, includeStartFile bool, limit int64, eachEntryFunc filer.ListEachEntryFunc) error {
+	//mc.RLock()
+	//defer mc.RUnlock()
+
+	if !mc.isCachedFn(dirPath) {
+		// if this request comes after renaming, it should be fine
+		glog.Warningf("unsynchronized dir: %v", dirPath)
+	}
+
+	_, err := mc.localStore.ListDirectoryEntries(ctx, dirPath, startFileName, includeStartFile, limit, func(entry *filer.Entry) bool {
+		mc.mapIdFromFilerToLocal(entry)
+		return eachEntryFunc(entry)
+	})
+	if err != nil {
+		return err
+	}
+	return err
+}
+
+func (mc *MetaCache) Shutdown() {
+	//mc.Lock()
+	//defer mc.Unlock()
+	mc.localStore.Shutdown()
+}
+
+func (mc *MetaCache) mapIdFromFilerToLocal(entry *filer.Entry) {
+	entry.Attr.Uid, entry.Attr.Gid = mc.uidGidMapper.FilerToLocal(entry.Attr.Uid, entry.Attr.Gid)
+}
+
+func (mc *MetaCache) Debug() {
+	if debuggable, ok := mc.localStore.(filer.Debuggable); ok {
+		println("start debugging")
+		debuggable.Debug(os.Stderr)
+	}
+}

weed/mount/meta_cache/meta_cache_init.go

@@ -0,0 +1,67 @@
+package meta_cache
+
+import (
+	"context"
+	"fmt"
+
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/chrislusf/seaweedfs/weed/util"
+)
+
+func EnsureVisited(mc *MetaCache, client filer_pb.FilerClient, dirPath util.FullPath) error {
+
+	for {
+
+		// the directory children are already cached
+		// so no need for this and upper directories
+		if mc.isCachedFn(dirPath) {
+			return nil
+		}
+
+		if err := doEnsureVisited(mc, client, dirPath); err != nil {
+			return err
+		}
+
+		// continue to parent directory
+		if dirPath != "/" {
+			parent, _ := dirPath.DirAndName()
+			dirPath = util.FullPath(parent)
+		} else {
+			break
+		}
+	}
+
+	return nil
+
+}
+
+func doEnsureVisited(mc *MetaCache, client filer_pb.FilerClient, path util.FullPath) error {
+
+	glog.V(4).Infof("ReadDirAllEntries %s ...", path)
+
+	err := util.Retry("ReadDirAllEntries", func() error {
+		return filer_pb.ReadDirAllEntries(client, path, "", func(pbEntry *filer_pb.Entry, isLast bool) error {
+			entry := filer.FromPbEntry(string(path), pbEntry)
+			if IsHiddenSystemEntry(string(path), entry.Name()) {
+				return nil
+			}
+			if err := mc.doInsertEntry(context.Background(), entry); err != nil {
+				glog.V(0).Infof("read %s: %v", entry.FullPath, err)
+				return err
+			}
+			return nil
+		})
+	})
+
+	if err != nil {
+		err = fmt.Errorf("list %s: %v", path, err)
+	}
+	mc.markCachedFn(path)
+	return err
+}
+
+func IsHiddenSystemEntry(dir, name string) bool {
+	return dir == "/" && (name == "topics" || name == "etc")
+}

weed/mount/meta_cache/meta_cache_subscribe.go

@@ -0,0 +1,68 @@
+package meta_cache
+
+import (
+	"context"
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/pb"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/chrislusf/seaweedfs/weed/util"
+)
+
+func SubscribeMetaEvents(mc *MetaCache, selfSignature int32, client filer_pb.FilerClient, dir string, lastTsNs int64) error {
+
+	processEventFn := func(resp *filer_pb.SubscribeMetadataResponse) error {
+		message := resp.EventNotification
+
+		for _, sig := range message.Signatures {
+			if sig == selfSignature && selfSignature != 0 {
+				return nil
+			}
+		}
+
+		dir := resp.Directory
+		var oldPath util.FullPath
+		var newEntry *filer.Entry
+		if message.OldEntry != nil {
+			oldPath = util.NewFullPath(dir, message.OldEntry.Name)
+			glog.V(4).Infof("deleting %v", oldPath)
+		}
+
+		if message.NewEntry != nil {
+			if message.NewParentPath != "" {
+				dir = message.NewParentPath
+			}
+			key := util.NewFullPath(dir, message.NewEntry.Name)
+			glog.V(4).Infof("creating %v", key)
+			newEntry = filer.FromPbEntry(dir, message.NewEntry)
+		}
+		err := mc.AtomicUpdateEntryFromFiler(context.Background(), oldPath, newEntry)
+		if err == nil {
+			if message.OldEntry != nil && message.NewEntry != nil {
+				oldKey := util.NewFullPath(resp.Directory, message.OldEntry.Name)
+				mc.invalidateFunc(oldKey, message.OldEntry)
+				if message.OldEntry.Name != message.NewEntry.Name {
+					newKey := util.NewFullPath(dir, message.NewEntry.Name)
+					mc.invalidateFunc(newKey, message.NewEntry)
+				}
+			} else if message.OldEntry == nil && message.NewEntry != nil {
+				// no need to invaalidate
+			} else if message.OldEntry != nil && message.NewEntry == nil {
+				oldKey := util.NewFullPath(resp.Directory, message.OldEntry.Name)
+				mc.invalidateFunc(oldKey, message.OldEntry)
+			}
+		}
+
+		return err
+
+	}
+
+	util.RetryForever("followMetaUpdates", func() error {
+		return pb.WithFilerClientFollowMetadata(client, "mount", selfSignature, dir, &lastTsNs, selfSignature, processEventFn, true)
+	}, func(err error) bool {
+		glog.Errorf("follow metadata updates: %v", err)
+		return true
+	})
+
+	return nil
+}

weed/mount/page_writer.go

@@ -0,0 +1,95 @@
+package mount
+
+import (
+	"github.com/chrislusf/seaweedfs/weed/filesys/page_writer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+)
+
+type PageWriter struct {
+	fh          *FileHandle
+	collection  string
+	replication string
+	chunkSize   int64
+
+	randomWriter page_writer.DirtyPages
+}
+
+var (
+	_ = page_writer.DirtyPages(&PageWriter{})
+)
+
+func newPageWriter(fh *FileHandle, chunkSize int64) *PageWriter {
+	pw := &PageWriter{
+		fh:           fh,
+		chunkSize:    chunkSize,
+		randomWriter: newMemoryChunkPages(fh, chunkSize),
+		// randomWriter: newTempFileDirtyPages(fh.f, chunkSize),
+	}
+	return pw
+}
+
+func (pw *PageWriter) AddPage(offset int64, data []byte) {
+
+	glog.V(4).Infof("%v AddPage [%d, %d)", pw.fh.fh, offset, offset+int64(len(data)))
+
+	chunkIndex := offset / pw.chunkSize
+	for i := chunkIndex; len(data) > 0; i++ {
+		writeSize := min(int64(len(data)), (i+1)*pw.chunkSize-offset)
+		pw.addToOneChunk(i, offset, data[:writeSize])
+		offset += writeSize
+		data = data[writeSize:]
+	}
+}
+
+func (pw *PageWriter) addToOneChunk(chunkIndex, offset int64, data []byte) {
+	pw.randomWriter.AddPage(offset, data)
+}
+
+func (pw *PageWriter) FlushData() error {
+	return pw.randomWriter.FlushData()
+}
+
+func (pw *PageWriter) ReadDirtyDataAt(data []byte, offset int64) (maxStop int64) {
+	glog.V(4).Infof("ReadDirtyDataAt %v [%d, %d)", pw.fh.fh, offset, offset+int64(len(data)))
+
+	chunkIndex := offset / pw.chunkSize
+	for i := chunkIndex; len(data) > 0; i++ {
+		readSize := min(int64(len(data)), (i+1)*pw.chunkSize-offset)
+
+		maxStop = pw.randomWriter.ReadDirtyDataAt(data[:readSize], offset)
+
+		offset += readSize
+		data = data[readSize:]
+	}
+
+	return
+}
+
+func (pw *PageWriter) GetStorageOptions() (collection, replication string) {
+	return pw.randomWriter.GetStorageOptions()
+}
+
+func (pw *PageWriter) LockForRead(startOffset, stopOffset int64) {
+	pw.randomWriter.LockForRead(startOffset, stopOffset)
+}
+
+func (pw *PageWriter) UnlockForRead(startOffset, stopOffset int64) {
+	pw.randomWriter.UnlockForRead(startOffset, stopOffset)
+}
+
+func (pw *PageWriter) Destroy() {
+	pw.randomWriter.Destroy()
+}
+
+func max(x, y int64) int64 {
+	if x > y {
+		return x
+	}
+	return y
+}
+func min(x, y int64) int64 {
+	if x < y {
+		return x
+	}
+	return y
+}

weed/mount/page_writer/chunk_interval_list.go

@@ -0,0 +1,115 @@
+package page_writer
+
+import "math"
+
+// ChunkWrittenInterval mark one written interval within one page chunk
+type ChunkWrittenInterval struct {
+	StartOffset int64
+	stopOffset  int64
+	prev        *ChunkWrittenInterval
+	next        *ChunkWrittenInterval
+}
+
+func (interval *ChunkWrittenInterval) Size() int64 {
+	return interval.stopOffset - interval.StartOffset
+}
+
+func (interval *ChunkWrittenInterval) isComplete(chunkSize int64) bool {
+	return interval.stopOffset-interval.StartOffset == chunkSize
+}
+
+// ChunkWrittenIntervalList mark written intervals within one page chunk
+type ChunkWrittenIntervalList struct {
+	head *ChunkWrittenInterval
+	tail *ChunkWrittenInterval
+}
+
+func newChunkWrittenIntervalList() *ChunkWrittenIntervalList {
+	list := &ChunkWrittenIntervalList{
+		head: &ChunkWrittenInterval{
+			StartOffset: -1,
+			stopOffset:  -1,
+		},
+		tail: &ChunkWrittenInterval{
+			StartOffset: math.MaxInt64,
+			stopOffset:  math.MaxInt64,
+		},
+	}
+	list.head.next = list.tail
+	list.tail.prev = list.head
+	return list
+}
+
+func (list *ChunkWrittenIntervalList) MarkWritten(startOffset, stopOffset int64) {
+	interval := &ChunkWrittenInterval{
+		StartOffset: startOffset,
+		stopOffset:  stopOffset,
+	}
+	list.addInterval(interval)
+}
+
+func (list *ChunkWrittenIntervalList) IsComplete(chunkSize int64) bool {
+	return list.size() == 1 && list.head.next.isComplete(chunkSize)
+}
+func (list *ChunkWrittenIntervalList) WrittenSize() (writtenByteCount int64) {
+	for t := list.head; t != nil; t = t.next {
+		writtenByteCount += t.Size()
+	}
+	return
+}
+
+func (list *ChunkWrittenIntervalList) addInterval(interval *ChunkWrittenInterval) {
+
+	p := list.head
+	for ; p.next != nil && p.next.StartOffset <= interval.StartOffset; p = p.next {
+	}
+	q := list.tail
+	for ; q.prev != nil && q.prev.stopOffset >= interval.stopOffset; q = q.prev {
+	}
+
+	if interval.StartOffset <= p.stopOffset && q.StartOffset <= interval.stopOffset {
+		// merge p and q together
+		p.stopOffset = q.stopOffset
+		unlinkNodesBetween(p, q.next)
+		return
+	}
+	if interval.StartOffset <= p.stopOffset {
+		// merge new interval into p
+		p.stopOffset = interval.stopOffset
+		unlinkNodesBetween(p, q)
+		return
+	}
+	if q.StartOffset <= interval.stopOffset {
+		// merge new interval into q
+		q.StartOffset = interval.StartOffset
+		unlinkNodesBetween(p, q)
+		return
+	}
+
+	// add the new interval between p and q
+	unlinkNodesBetween(p, q)
+	p.next = interval
+	interval.prev = p
+	q.prev = interval
+	interval.next = q
+
+}
+
+// unlinkNodesBetween remove all nodes after start and before stop, exclusive
+func unlinkNodesBetween(start *ChunkWrittenInterval, stop *ChunkWrittenInterval) {
+	if start.next == stop {
+		return
+	}
+	start.next.prev = nil
+	start.next = stop
+	stop.prev.next = nil
+	stop.prev = start
+}
+
+func (list *ChunkWrittenIntervalList) size() int {
+	var count int
+	for t := list.head; t != nil; t = t.next {
+		count++
+	}
+	return count - 2
+}

weed/mount/page_writer/chunk_interval_list_test.go

@@ -0,0 +1,49 @@
+package page_writer
+
+import (
+	"github.com/stretchr/testify/assert"
+	"testing"
+)
+
+func Test_PageChunkWrittenIntervalList(t *testing.T) {
+	list := newChunkWrittenIntervalList()
+
+	assert.Equal(t, 0, list.size(), "empty list")
+
+	list.MarkWritten(0, 5)
+	assert.Equal(t, 1, list.size(), "one interval")
+
+	list.MarkWritten(0, 5)
+	assert.Equal(t, 1, list.size(), "duplicated interval2")
+
+	list.MarkWritten(95, 100)
+	assert.Equal(t, 2, list.size(), "two intervals")
+
+	list.MarkWritten(50, 60)
+	assert.Equal(t, 3, list.size(), "three intervals")
+
+	list.MarkWritten(50, 55)
+	assert.Equal(t, 3, list.size(), "three intervals merge")
+
+	list.MarkWritten(40, 50)
+	assert.Equal(t, 3, list.size(), "three intervals grow forward")
+
+	list.MarkWritten(50, 65)
+	assert.Equal(t, 3, list.size(), "three intervals grow backward")
+
+	list.MarkWritten(70, 80)
+	assert.Equal(t, 4, list.size(), "four intervals")
+
+	list.MarkWritten(60, 70)
+	assert.Equal(t, 3, list.size(), "three intervals merged")
+
+	list.MarkWritten(59, 71)
+	assert.Equal(t, 3, list.size(), "covered three intervals")
+
+	list.MarkWritten(5, 59)
+	assert.Equal(t, 2, list.size(), "covered two intervals")
+
+	list.MarkWritten(70, 99)
+	assert.Equal(t, 1, list.size(), "covered one intervals")
+
+}

weed/mount/page_writer/dirty_pages.go

@@ -0,0 +1,30 @@
+package page_writer
+
+type DirtyPages interface {
+	AddPage(offset int64, data []byte)
+	FlushData() error
+	ReadDirtyDataAt(data []byte, startOffset int64) (maxStop int64)
+	GetStorageOptions() (collection, replication string)
+	Destroy()
+	LockForRead(startOffset, stopOffset int64)
+	UnlockForRead(startOffset, stopOffset int64)
+}
+
+func max(x, y int64) int64 {
+	if x > y {
+		return x
+	}
+	return y
+}
+func min(x, y int64) int64 {
+	if x < y {
+		return x
+	}
+	return y
+}
+func minInt(x, y int) int {
+	if x < y {
+		return x
+	}
+	return y
+}

weed/mount/page_writer/page_chunk.go

@@ -0,0 +1,16 @@
+package page_writer
+
+import (
+	"io"
+)
+
+type SaveToStorageFunc func(reader io.Reader, offset int64, size int64, cleanupFn func())
+
+type PageChunk interface {
+	FreeResource()
+	WriteDataAt(src []byte, offset int64) (n int)
+	ReadDataAt(p []byte, off int64) (maxStop int64)
+	IsComplete() bool
+	WrittenSize() int64
+	SaveContent(saveFn SaveToStorageFunc)
+}

weed/mount/page_writer/page_chunk_mem.go

@@ -0,0 +1,69 @@
+package page_writer
+
+import (
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"github.com/chrislusf/seaweedfs/weed/util/mem"
+)
+
+var (
+	_ = PageChunk(&MemChunk{})
+)
+
+type MemChunk struct {
+	buf             []byte
+	usage           *ChunkWrittenIntervalList
+	chunkSize       int64
+	logicChunkIndex LogicChunkIndex
+}
+
+func NewMemChunk(logicChunkIndex LogicChunkIndex, chunkSize int64) *MemChunk {
+	return &MemChunk{
+		logicChunkIndex: logicChunkIndex,
+		chunkSize:       chunkSize,
+		buf:             mem.Allocate(int(chunkSize)),
+		usage:           newChunkWrittenIntervalList(),
+	}
+}
+
+func (mc *MemChunk) FreeResource() {
+	mem.Free(mc.buf)
+}
+
+func (mc *MemChunk) WriteDataAt(src []byte, offset int64) (n int) {
+	innerOffset := offset % mc.chunkSize
+	n = copy(mc.buf[innerOffset:], src)
+	mc.usage.MarkWritten(innerOffset, innerOffset+int64(n))
+	return
+}
+
+func (mc *MemChunk) ReadDataAt(p []byte, off int64) (maxStop int64) {
+	memChunkBaseOffset := int64(mc.logicChunkIndex) * mc.chunkSize
+	for t := mc.usage.head.next; t != mc.usage.tail; t = t.next {
+		logicStart := max(off, int64(mc.logicChunkIndex)*mc.chunkSize+t.StartOffset)
+		logicStop := min(off+int64(len(p)), memChunkBaseOffset+t.stopOffset)
+		if logicStart < logicStop {
+			copy(p[logicStart-off:logicStop-off], mc.buf[logicStart-memChunkBaseOffset:logicStop-memChunkBaseOffset])
+			maxStop = max(maxStop, logicStop)
+		}
+	}
+	return
+}
+
+func (mc *MemChunk) IsComplete() bool {
+	return mc.usage.IsComplete(mc.chunkSize)
+}
+
+func (mc *MemChunk) WrittenSize() int64 {
+	return mc.usage.WrittenSize()
+}
+
+func (mc *MemChunk) SaveContent(saveFn SaveToStorageFunc) {
+	if saveFn == nil {
+		return
+	}
+	for t := mc.usage.head.next; t != mc.usage.tail; t = t.next {
+		reader := util.NewBytesReader(mc.buf[t.StartOffset:t.stopOffset])
+		saveFn(reader, int64(mc.logicChunkIndex)*mc.chunkSize+t.StartOffset, t.Size(), func() {
+		})
+	}
+}

weed/mount/page_writer/page_chunk_swapfile.go

@@ -0,0 +1,121 @@
+package page_writer
+
+import (
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"github.com/chrislusf/seaweedfs/weed/util/mem"
+	"os"
+)
+
+var (
+	_ = PageChunk(&SwapFileChunk{})
+)
+
+type ActualChunkIndex int
+
+type SwapFile struct {
+	dir                     string
+	file                    *os.File
+	logicToActualChunkIndex map[LogicChunkIndex]ActualChunkIndex
+	chunkSize               int64
+}
+
+type SwapFileChunk struct {
+	swapfile         *SwapFile
+	usage            *ChunkWrittenIntervalList
+	logicChunkIndex  LogicChunkIndex
+	actualChunkIndex ActualChunkIndex
+}
+
+func NewSwapFile(dir string, chunkSize int64) *SwapFile {
+	return &SwapFile{
+		dir:                     dir,
+		file:                    nil,
+		logicToActualChunkIndex: make(map[LogicChunkIndex]ActualChunkIndex),
+		chunkSize:               chunkSize,
+	}
+}
+func (sf *SwapFile) FreeResource() {
+	if sf.file != nil {
+		sf.file.Close()
+		os.Remove(sf.file.Name())
+	}
+}
+
+func (sf *SwapFile) NewTempFileChunk(logicChunkIndex LogicChunkIndex) (tc *SwapFileChunk) {
+	if sf.file == nil {
+		var err error
+		sf.file, err = os.CreateTemp(sf.dir, "")
+		if err != nil {
+			glog.Errorf("create swap file: %v", err)
+			return nil
+		}
+	}
+	actualChunkIndex, found := sf.logicToActualChunkIndex[logicChunkIndex]
+	if !found {
+		actualChunkIndex = ActualChunkIndex(len(sf.logicToActualChunkIndex))
+		sf.logicToActualChunkIndex[logicChunkIndex] = actualChunkIndex
+	}
+
+	return &SwapFileChunk{
+		swapfile:         sf,
+		usage:            newChunkWrittenIntervalList(),
+		logicChunkIndex:  logicChunkIndex,
+		actualChunkIndex: actualChunkIndex,
+	}
+}
+
+func (sc *SwapFileChunk) FreeResource() {
+}
+
+func (sc *SwapFileChunk) WriteDataAt(src []byte, offset int64) (n int) {
+	innerOffset := offset % sc.swapfile.chunkSize
+	var err error
+	n, err = sc.swapfile.file.WriteAt(src, int64(sc.actualChunkIndex)*sc.swapfile.chunkSize+innerOffset)
+	if err == nil {
+		sc.usage.MarkWritten(innerOffset, innerOffset+int64(n))
+	} else {
+		glog.Errorf("failed to write swap file %s: %v", sc.swapfile.file.Name(), err)
+	}
+	return
+}
+
+func (sc *SwapFileChunk) ReadDataAt(p []byte, off int64) (maxStop int64) {
+	chunkStartOffset := int64(sc.logicChunkIndex) * sc.swapfile.chunkSize
+	for t := sc.usage.head.next; t != sc.usage.tail; t = t.next {
+		logicStart := max(off, chunkStartOffset+t.StartOffset)
+		logicStop := min(off+int64(len(p)), chunkStartOffset+t.stopOffset)
+		if logicStart < logicStop {
+			actualStart := logicStart - chunkStartOffset + int64(sc.actualChunkIndex)*sc.swapfile.chunkSize
+			if _, err := sc.swapfile.file.ReadAt(p[logicStart-off:logicStop-off], actualStart); err != nil {
+				glog.Errorf("failed to reading swap file %s: %v", sc.swapfile.file.Name(), err)
+				break
+			}
+			maxStop = max(maxStop, logicStop)
+		}
+	}
+	return
+}
+
+func (sc *SwapFileChunk) IsComplete() bool {
+	return sc.usage.IsComplete(sc.swapfile.chunkSize)
+}
+
+func (sc *SwapFileChunk) WrittenSize() int64 {
+	return sc.usage.WrittenSize()
+}
+
+func (sc *SwapFileChunk) SaveContent(saveFn SaveToStorageFunc) {
+	if saveFn == nil {
+		return
+	}
+	for t := sc.usage.head.next; t != sc.usage.tail; t = t.next {
+		data := mem.Allocate(int(t.Size()))
+		sc.swapfile.file.ReadAt(data, t.StartOffset+int64(sc.actualChunkIndex)*sc.swapfile.chunkSize)
+		reader := util.NewBytesReader(data)
+		saveFn(reader, int64(sc.logicChunkIndex)*sc.swapfile.chunkSize+t.StartOffset, t.Size(), func() {
+		})
+		mem.Free(data)
+	}
+	sc.usage = newChunkWrittenIntervalList()
+}

weed/mount/page_writer/upload_pipeline.go

@@ -0,0 +1,182 @@
+package page_writer
+
+import (
+	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"sync"
+	"sync/atomic"
+	"time"
+)
+
+type LogicChunkIndex int
+
+type UploadPipeline struct {
+	filepath             util.FullPath
+	ChunkSize            int64
+	writableChunks       map[LogicChunkIndex]PageChunk
+	writableChunksLock   sync.Mutex
+	sealedChunks         map[LogicChunkIndex]*SealedChunk
+	sealedChunksLock     sync.Mutex
+	uploaders            *util.LimitedConcurrentExecutor
+	uploaderCount        int32
+	uploaderCountCond    *sync.Cond
+	saveToStorageFn      SaveToStorageFunc
+	activeReadChunks     map[LogicChunkIndex]int
+	activeReadChunksLock sync.Mutex
+	bufferChunkLimit     int
+}
+
+type SealedChunk struct {
+	chunk            PageChunk
+	referenceCounter int // track uploading or reading processes
+}
+
+func (sc *SealedChunk) FreeReference(messageOnFree string) {
+	sc.referenceCounter--
+	if sc.referenceCounter == 0 {
+		glog.V(4).Infof("Free sealed chunk: %s", messageOnFree)
+		sc.chunk.FreeResource()
+	}
+}
+
+func NewUploadPipeline(writers *util.LimitedConcurrentExecutor, chunkSize int64, saveToStorageFn SaveToStorageFunc, bufferChunkLimit int) *UploadPipeline {
+	return &UploadPipeline{
+		ChunkSize:         chunkSize,
+		writableChunks:    make(map[LogicChunkIndex]PageChunk),
+		sealedChunks:      make(map[LogicChunkIndex]*SealedChunk),
+		uploaders:         writers,
+		uploaderCountCond: sync.NewCond(&sync.Mutex{}),
+		saveToStorageFn:   saveToStorageFn,
+		activeReadChunks:  make(map[LogicChunkIndex]int),
+		bufferChunkLimit:  bufferChunkLimit,
+	}
+}
+
+func (up *UploadPipeline) SaveDataAt(p []byte, off int64) (n int) {
+	up.writableChunksLock.Lock()
+	defer up.writableChunksLock.Unlock()
+
+	logicChunkIndex := LogicChunkIndex(off / up.ChunkSize)
+
+	memChunk, found := up.writableChunks[logicChunkIndex]
+	if !found {
+		if len(up.writableChunks) < up.bufferChunkLimit {
+			memChunk = NewMemChunk(logicChunkIndex, up.ChunkSize)
+		} else {
+			fullestChunkIndex, fullness := LogicChunkIndex(-1), int64(0)
+			for lci, mc := range up.writableChunks {
+				chunkFullness := mc.WrittenSize()
+				if fullness < chunkFullness {
+					fullestChunkIndex = lci
+					fullness = chunkFullness
+				}
+			}
+			up.moveToSealed(up.writableChunks[fullestChunkIndex], fullestChunkIndex)
+			delete(up.writableChunks, fullestChunkIndex)
+			fmt.Printf("flush chunk %d with %d bytes written", logicChunkIndex, fullness)
+			memChunk = NewMemChunk(logicChunkIndex, up.ChunkSize)
+		}
+		up.writableChunks[logicChunkIndex] = memChunk
+	}
+	n = memChunk.WriteDataAt(p, off)
+	up.maybeMoveToSealed(memChunk, logicChunkIndex)
+
+	return
+}
+
+func (up *UploadPipeline) MaybeReadDataAt(p []byte, off int64) (maxStop int64) {
+	logicChunkIndex := LogicChunkIndex(off / up.ChunkSize)
+
+	// read from sealed chunks first
+	up.sealedChunksLock.Lock()
+	sealedChunk, found := up.sealedChunks[logicChunkIndex]
+	if found {
+		sealedChunk.referenceCounter++
+	}
+	up.sealedChunksLock.Unlock()
+	if found {
+		maxStop = sealedChunk.chunk.ReadDataAt(p, off)
+		glog.V(4).Infof("%s read sealed memchunk [%d,%d)", up.filepath, off, maxStop)
+		sealedChunk.FreeReference(fmt.Sprintf("%s finish reading chunk %d", up.filepath, logicChunkIndex))
+	}
+
+	// read from writable chunks last
+	up.writableChunksLock.Lock()
+	defer up.writableChunksLock.Unlock()
+	writableChunk, found := up.writableChunks[logicChunkIndex]
+	if !found {
+		return
+	}
+	writableMaxStop := writableChunk.ReadDataAt(p, off)
+	glog.V(4).Infof("%s read writable memchunk [%d,%d)", up.filepath, off, writableMaxStop)
+	maxStop = max(maxStop, writableMaxStop)
+
+	return
+}
+
+func (up *UploadPipeline) FlushAll() {
+	up.writableChunksLock.Lock()
+	defer up.writableChunksLock.Unlock()
+
+	for logicChunkIndex, memChunk := range up.writableChunks {
+		up.moveToSealed(memChunk, logicChunkIndex)
+	}
+
+	up.waitForCurrentWritersToComplete()
+}
+
+func (up *UploadPipeline) maybeMoveToSealed(memChunk PageChunk, logicChunkIndex LogicChunkIndex) {
+	if memChunk.IsComplete() {
+		up.moveToSealed(memChunk, logicChunkIndex)
+	}
+}
+
+func (up *UploadPipeline) moveToSealed(memChunk PageChunk, logicChunkIndex LogicChunkIndex) {
+	atomic.AddInt32(&up.uploaderCount, 1)
+	glog.V(4).Infof("%s uploaderCount %d ++> %d", up.filepath, up.uploaderCount-1, up.uploaderCount)
+
+	up.sealedChunksLock.Lock()
+
+	if oldMemChunk, found := up.sealedChunks[logicChunkIndex]; found {
+		oldMemChunk.FreeReference(fmt.Sprintf("%s replace chunk %d", up.filepath, logicChunkIndex))
+	}
+	sealedChunk := &SealedChunk{
+		chunk:            memChunk,
+		referenceCounter: 1, // default 1 is for uploading process
+	}
+	up.sealedChunks[logicChunkIndex] = sealedChunk
+	delete(up.writableChunks, logicChunkIndex)
+
+	up.sealedChunksLock.Unlock()
+
+	up.uploaders.Execute(func() {
+		// first add to the file chunks
+		sealedChunk.chunk.SaveContent(up.saveToStorageFn)
+
+		// notify waiting process
+		atomic.AddInt32(&up.uploaderCount, -1)
+		glog.V(4).Infof("%s uploaderCount %d --> %d", up.filepath, up.uploaderCount+1, up.uploaderCount)
+		// Lock and Unlock are not required,
+		// but it may signal multiple times during one wakeup,
+		// and the waiting goroutine may miss some of them!
+		up.uploaderCountCond.L.Lock()
+		up.uploaderCountCond.Broadcast()
+		up.uploaderCountCond.L.Unlock()
+
+		// wait for readers
+		for up.IsLocked(logicChunkIndex) {
+			time.Sleep(59 * time.Millisecond)
+		}
+
+		// then remove from sealed chunks
+		up.sealedChunksLock.Lock()
+		defer up.sealedChunksLock.Unlock()
+		delete(up.sealedChunks, logicChunkIndex)
+		sealedChunk.FreeReference(fmt.Sprintf("%s finished uploading chunk %d", up.filepath, logicChunkIndex))
+
+	})
+}
+
+func (up *UploadPipeline) Shutdown() {
+}

weed/mount/page_writer/upload_pipeline_lock.go

@@ -0,0 +1,63 @@
+package page_writer
+
+import (
+	"sync/atomic"
+)
+
+func (up *UploadPipeline) LockForRead(startOffset, stopOffset int64) {
+	startLogicChunkIndex := LogicChunkIndex(startOffset / up.ChunkSize)
+	stopLogicChunkIndex := LogicChunkIndex(stopOffset / up.ChunkSize)
+	if stopOffset%up.ChunkSize > 0 {
+		stopLogicChunkIndex += 1
+	}
+	up.activeReadChunksLock.Lock()
+	defer up.activeReadChunksLock.Unlock()
+	for i := startLogicChunkIndex; i < stopLogicChunkIndex; i++ {
+		if count, found := up.activeReadChunks[i]; found {
+			up.activeReadChunks[i] = count + 1
+		} else {
+			up.activeReadChunks[i] = 1
+		}
+	}
+}
+
+func (up *UploadPipeline) UnlockForRead(startOffset, stopOffset int64) {
+	startLogicChunkIndex := LogicChunkIndex(startOffset / up.ChunkSize)
+	stopLogicChunkIndex := LogicChunkIndex(stopOffset / up.ChunkSize)
+	if stopOffset%up.ChunkSize > 0 {
+		stopLogicChunkIndex += 1
+	}
+	up.activeReadChunksLock.Lock()
+	defer up.activeReadChunksLock.Unlock()
+	for i := startLogicChunkIndex; i < stopLogicChunkIndex; i++ {
+		if count, found := up.activeReadChunks[i]; found {
+			if count == 1 {
+				delete(up.activeReadChunks, i)
+			} else {
+				up.activeReadChunks[i] = count - 1
+			}
+		}
+	}
+}
+
+func (up *UploadPipeline) IsLocked(logicChunkIndex LogicChunkIndex) bool {
+	up.activeReadChunksLock.Lock()
+	defer up.activeReadChunksLock.Unlock()
+	if count, found := up.activeReadChunks[logicChunkIndex]; found {
+		return count > 0
+	}
+	return false
+}
+
+func (up *UploadPipeline) waitForCurrentWritersToComplete() {
+	up.uploaderCountCond.L.Lock()
+	t := int32(100)
+	for {
+		t = atomic.LoadInt32(&up.uploaderCount)
+		if t <= 0 {
+			break
+		}
+		up.uploaderCountCond.Wait()
+	}
+	up.uploaderCountCond.L.Unlock()
+}

weed/mount/page_writer/upload_pipeline_test.go

@@ -0,0 +1,47 @@
+package page_writer
+
+import (
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"testing"
+)
+
+func TestUploadPipeline(t *testing.T) {
+
+	uploadPipeline := NewUploadPipeline(nil, 2*1024*1024, nil, 16)
+
+	writeRange(uploadPipeline, 0, 131072)
+	writeRange(uploadPipeline, 131072, 262144)
+	writeRange(uploadPipeline, 262144, 1025536)
+
+	confirmRange(t, uploadPipeline, 0, 1025536)
+
+	writeRange(uploadPipeline, 1025536, 1296896)
+
+	confirmRange(t, uploadPipeline, 1025536, 1296896)
+
+	writeRange(uploadPipeline, 1296896, 2162688)
+
+	confirmRange(t, uploadPipeline, 1296896, 2162688)
+
+	confirmRange(t, uploadPipeline, 1296896, 2162688)
+}
+
+// startOff and stopOff must be divided by 4
+func writeRange(uploadPipeline *UploadPipeline, startOff, stopOff int64) {
+	p := make([]byte, 4)
+	for i := startOff / 4; i < stopOff/4; i += 4 {
+		util.Uint32toBytes(p, uint32(i))
+		uploadPipeline.SaveDataAt(p, i)
+	}
+}
+
+func confirmRange(t *testing.T, uploadPipeline *UploadPipeline, startOff, stopOff int64) {
+	p := make([]byte, 4)
+	for i := startOff; i < stopOff/4; i += 4 {
+		uploadPipeline.MaybeReadDataAt(p, i)
+		x := util.BytesToUint32(p)
+		if x != uint32(i) {
+			t.Errorf("expecting %d found %d at offset [%d,%d)", i, x, i, i+4)
+		}
+	}
+}

weed/mount/unmount/unmount.go

@@ -0,0 +1,6 @@
+package unmount
+
+// Unmount tries to unmount the filesystem mounted at dir.
+func Unmount(dir string) error {
+	return unmount(dir)
+}

weed/mount/unmount/unmount_linux.go

@@ -0,0 +1,21 @@
+package unmount
+
+import (
+	"bytes"
+	"errors"
+	"os/exec"
+)
+
+func unmount(dir string) error {
+	cmd := exec.Command("fusermount", "-u", dir)
+	output, err := cmd.CombinedOutput()
+	if err != nil {
+		if len(output) > 0 {
+			output = bytes.TrimRight(output, "\n")
+			msg := err.Error() + ": " + string(output)
+			err = errors.New(msg)
+		}
+		return err
+	}
+	return nil
+}

weed/mount/unmount/unmount_std.go

@@ -0,0 +1,18 @@
+//go:build !linux
+// +build !linux
+
+package unmount
+
+import (
+	"os"
+	"syscall"
+)
+
+func unmount(dir string) error {
+	err := syscall.Unmount(dir, 0)
+	if err != nil {
+		err = &os.PathError{Op: "unmount", Path: dir, Err: err}
+		return err
+	}
+	return nil
+}

weed/mount/weedfs.go

@@ -0,0 +1,187 @@
+package mount
+
+import (
+	"context"
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/mount/meta_cache"
+	"github.com/chrislusf/seaweedfs/weed/pb"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/chrislusf/seaweedfs/weed/storage/types"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"github.com/chrislusf/seaweedfs/weed/util/chunk_cache"
+	"github.com/chrislusf/seaweedfs/weed/util/grace"
+	"github.com/chrislusf/seaweedfs/weed/wdclient"
+	"github.com/hanwen/go-fuse/v2/fuse"
+	"google.golang.org/grpc"
+	"math/rand"
+	"os"
+	"path"
+	"path/filepath"
+	"time"
+
+	"github.com/hanwen/go-fuse/v2/fs"
+)
+
+type Option struct {
+	MountDirectory     string
+	FilerAddresses     []pb.ServerAddress
+	filerIndex         int
+	GrpcDialOption     grpc.DialOption
+	FilerMountRootPath string
+	Collection         string
+	Replication        string
+	TtlSec             int32
+	DiskType           types.DiskType
+	ChunkSizeLimit     int64
+	ConcurrentWriters  int
+	CacheDir           string
+	CacheSizeMB        int64
+	DataCenter         string
+	Umask              os.FileMode
+
+	MountUid         uint32
+	MountGid         uint32
+	MountMode        os.FileMode
+	MountCtime       time.Time
+	MountMtime       time.Time
+	MountParentInode uint64
+
+	VolumeServerAccess string // how to access volume servers
+	Cipher             bool   // whether encrypt data on volume server
+	UidGidMapper       *meta_cache.UidGidMapper
+
+	uniqueCacheDir         string
+	uniqueCacheTempPageDir string
+}
+
+type WFS struct {
+	// follow https://github.com/hanwen/go-fuse/blob/master/fuse/api.go
+	fuse.RawFileSystem
+	fs.Inode
+	option            *Option
+	metaCache         *meta_cache.MetaCache
+	stats             statsCache
+	root              Directory
+	chunkCache        *chunk_cache.TieredChunkCache
+	signature         int32
+	concurrentWriters *util.LimitedConcurrentExecutor
+	inodeToPath       *InodeToPath
+	fhmap             *FileHandleToInode
+}
+
+func NewSeaweedFileSystem(option *Option) *WFS {
+	wfs := &WFS{
+		RawFileSystem: fuse.NewDefaultRawFileSystem(),
+		option:        option,
+		signature:     util.RandomInt32(),
+		inodeToPath:   NewInodeToPath(),
+		fhmap:         NewFileHandleToInode(),
+	}
+
+	wfs.root = Directory{
+		name:   "/",
+		wfs:    wfs,
+		entry:  nil,
+		parent: nil,
+	}
+
+	wfs.option.filerIndex = rand.Intn(len(option.FilerAddresses))
+	wfs.option.setupUniqueCacheDirectory()
+	if option.CacheSizeMB > 0 {
+		wfs.chunkCache = chunk_cache.NewTieredChunkCache(256, option.getUniqueCacheDir(), option.CacheSizeMB, 1024*1024)
+	}
+
+	wfs.metaCache = meta_cache.NewMetaCache(path.Join(option.getUniqueCacheDir(), "meta"), option.UidGidMapper, func(path util.FullPath) {
+		wfs.inodeToPath.MarkChildrenCached(path)
+	}, func(path util.FullPath) bool {
+		return wfs.inodeToPath.IsChildrenCached(path)
+	}, func(filePath util.FullPath, entry *filer_pb.Entry) {
+	})
+	grace.OnInterrupt(func() {
+		wfs.metaCache.Shutdown()
+	})
+
+	if wfs.option.ConcurrentWriters > 0 {
+		wfs.concurrentWriters = util.NewLimitedConcurrentExecutor(wfs.option.ConcurrentWriters)
+	}
+	return wfs
+}
+
+func (wfs *WFS) StartBackgroundTasks() {
+	startTime := time.Now()
+	go meta_cache.SubscribeMetaEvents(wfs.metaCache, wfs.signature, wfs, wfs.option.FilerMountRootPath, startTime.UnixNano())
+}
+
+func (wfs *WFS) Root() *Directory {
+	return &wfs.root
+}
+
+func (wfs *WFS) String() string {
+	return "seaweedfs"
+}
+
+func (wfs *WFS) maybeReadEntry(inode uint64) (path util.FullPath, fh *FileHandle, entry *filer_pb.Entry, status fuse.Status) {
+	path = wfs.inodeToPath.GetPath(inode)
+	var found bool
+	if fh, found = wfs.fhmap.FindFileHandle(inode); found {
+		return path, fh, fh.entry, fuse.OK
+	}
+	entry, status = wfs.maybeLoadEntry(path)
+	return
+}
+
+func (wfs *WFS) maybeLoadEntry(fullpath util.FullPath) (*filer_pb.Entry, fuse.Status) {
+
+	// glog.V(3).Infof("read entry cache miss %s", fullpath)
+	dir, name := fullpath.DirAndName()
+
+	// return a valid entry for the mount root
+	if string(fullpath) == wfs.option.FilerMountRootPath {
+		return &filer_pb.Entry{
+			Name:        name,
+			IsDirectory: true,
+			Attributes: &filer_pb.FuseAttributes{
+				Mtime:    wfs.option.MountMtime.Unix(),
+				FileMode: uint32(wfs.option.MountMode),
+				Uid:      wfs.option.MountUid,
+				Gid:      wfs.option.MountGid,
+				Crtime:   wfs.option.MountCtime.Unix(),
+			},
+		}, fuse.OK
+	}
+
+	// read from async meta cache
+	meta_cache.EnsureVisited(wfs.metaCache, wfs, util.FullPath(dir))
+	cachedEntry, cacheErr := wfs.metaCache.FindEntry(context.Background(), fullpath)
+	if cacheErr == filer_pb.ErrNotFound {
+		return nil, fuse.ENOENT
+	}
+	return cachedEntry.ToProtoEntry(), fuse.OK
+}
+
+func (wfs *WFS) LookupFn() wdclient.LookupFileIdFunctionType {
+	if wfs.option.VolumeServerAccess == "filerProxy" {
+		return func(fileId string) (targetUrls []string, err error) {
+			return []string{"http://" + wfs.getCurrentFiler().ToHttpAddress() + "/?proxyChunkId=" + fileId}, nil
+		}
+	}
+	return filer.LookupFn(wfs)
+}
+
+func (wfs *WFS) getCurrentFiler() pb.ServerAddress {
+	return wfs.option.FilerAddresses[wfs.option.filerIndex]
+}
+
+func (option *Option) setupUniqueCacheDirectory() {
+	cacheUniqueId := util.Md5String([]byte(option.MountDirectory + string(option.FilerAddresses[0]) + option.FilerMountRootPath + util.Version()))[0:8]
+	option.uniqueCacheDir = path.Join(option.CacheDir, cacheUniqueId)
+	option.uniqueCacheTempPageDir = filepath.Join(option.uniqueCacheDir, "sw")
+	os.MkdirAll(option.uniqueCacheTempPageDir, os.FileMode(0777)&^option.Umask)
+}
+
+func (option *Option) getTempFilePageDir() string {
+	return option.uniqueCacheTempPageDir
+}
+func (option *Option) getUniqueCacheDir() string {
+	return option.uniqueCacheDir
+}

weed/mount/weedfs_attr.go

@@ -0,0 +1,208 @@
+package mount
+
+import (
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/hanwen/go-fuse/v2/fuse"
+	"os"
+	"syscall"
+	"time"
+)
+
+func (wfs *WFS) GetAttr(cancel <-chan struct{}, input *fuse.GetAttrIn, out *fuse.AttrOut) (code fuse.Status) {
+	if input.NodeId == 1 {
+		wfs.setRootAttr(out)
+		return fuse.OK
+	}
+
+	_, _, entry, status := wfs.maybeReadEntry(input.NodeId)
+	if status != fuse.OK {
+		return status
+	}
+	out.AttrValid = 1
+	wfs.setAttrByPbEntry(&out.Attr, input.NodeId, entry)
+
+	return fuse.OK
+}
+
+func (wfs *WFS) SetAttr(cancel <-chan struct{}, input *fuse.SetAttrIn, out *fuse.AttrOut) (code fuse.Status) {
+
+	path, fh, entry, status := wfs.maybeReadEntry(input.NodeId)
+	if status != fuse.OK {
+		return status
+	}
+
+	if size, ok := input.GetSize(); ok {
+		glog.V(4).Infof("%v setattr set size=%v chunks=%d", path, size, len(entry.Chunks))
+		if size < filer.FileSize(entry) {
+			// fmt.Printf("truncate %v \n", fullPath)
+			var chunks []*filer_pb.FileChunk
+			var truncatedChunks []*filer_pb.FileChunk
+			for _, chunk := range entry.Chunks {
+				int64Size := int64(chunk.Size)
+				if chunk.Offset+int64Size > int64(size) {
+					// this chunk is truncated
+					int64Size = int64(size) - chunk.Offset
+					if int64Size > 0 {
+						chunks = append(chunks, chunk)
+						glog.V(4).Infof("truncated chunk %+v from %d to %d\n", chunk.GetFileIdString(), chunk.Size, int64Size)
+						chunk.Size = uint64(int64Size)
+					} else {
+						glog.V(4).Infof("truncated whole chunk %+v\n", chunk.GetFileIdString())
+						truncatedChunks = append(truncatedChunks, chunk)
+					}
+				}
+			}
+			// set the new chunks and reset entry cache
+			entry.Chunks = chunks
+			if fh != nil {
+				fh.entryViewCache = nil
+			}
+		}
+		entry.Attributes.Mtime = time.Now().Unix()
+		entry.Attributes.FileSize = size
+
+	}
+
+	if mode, ok := input.GetMode(); ok {
+		entry.Attributes.FileMode = uint32(mode)
+	}
+
+	if uid, ok := input.GetUID(); ok {
+		entry.Attributes.Uid = uid
+	}
+
+	if gid, ok := input.GetGID(); ok {
+		entry.Attributes.Gid = gid
+	}
+
+	if mtime, ok := input.GetMTime(); ok {
+		entry.Attributes.Mtime = mtime.Unix()
+	}
+
+	entry.Attributes.Mtime = time.Now().Unix()
+	out.AttrValid = 1
+	wfs.setAttrByPbEntry(&out.Attr, input.NodeId, entry)
+
+	if fh != nil {
+		fh.dirtyMetadata = true
+		return fuse.OK
+	}
+
+	return wfs.saveEntry(path, entry)
+
+}
+
+func (wfs *WFS) setRootAttr(out *fuse.AttrOut) {
+	now := uint64(time.Now().Unix())
+	out.AttrValid = 119
+	out.Ino = 1
+	setBlksize(&out.Attr, blockSize)
+	out.Uid = wfs.option.MountUid
+	out.Gid = wfs.option.MountGid
+	out.Mtime = now
+	out.Ctime = now
+	out.Atime = now
+	out.Mode = toSystemType(os.ModeDir) | uint32(wfs.option.MountMode)
+	out.Nlink = 1
+}
+
+func (wfs *WFS) setAttrByPbEntry(out *fuse.Attr, inode uint64, entry *filer_pb.Entry) {
+	out.Ino = inode
+	out.Size = filer.FileSize(entry)
+	out.Blocks = (out.Size + blockSize - 1) / blockSize
+	setBlksize(out, blockSize)
+	out.Mtime = uint64(entry.Attributes.Mtime)
+	out.Ctime = uint64(entry.Attributes.Mtime)
+	out.Atime = uint64(entry.Attributes.Mtime)
+	out.Mode = toSystemMode(os.FileMode(entry.Attributes.FileMode))
+	if entry.HardLinkCounter > 0 {
+		out.Nlink = uint32(entry.HardLinkCounter)
+	} else {
+		out.Nlink = 1
+	}
+	out.Uid = entry.Attributes.Uid
+	out.Gid = entry.Attributes.Gid
+}
+
+func (wfs *WFS) setAttrByFilerEntry(out *fuse.Attr, inode uint64, entry *filer.Entry) {
+	out.Ino = inode
+	out.Size = entry.FileSize
+	out.Blocks = (out.Size + blockSize - 1) / blockSize
+	setBlksize(out, blockSize)
+	out.Atime = uint64(entry.Attr.Mtime.Unix())
+	out.Mtime = uint64(entry.Attr.Mtime.Unix())
+	out.Ctime = uint64(entry.Attr.Mtime.Unix())
+	out.Crtime_ = uint64(entry.Attr.Crtime.Unix())
+	out.Mode = toSystemMode(entry.Attr.Mode)
+	if entry.HardLinkCounter > 0 {
+		out.Nlink = uint32(entry.HardLinkCounter)
+	} else {
+		out.Nlink = 1
+	}
+	out.Uid = entry.Attr.Uid
+	out.Gid = entry.Attr.Gid
+}
+
+func (wfs *WFS) outputPbEntry(out *fuse.EntryOut, inode uint64, entry *filer_pb.Entry) {
+	out.NodeId = inode
+	out.Generation = 1
+	out.EntryValid = 1
+	out.AttrValid = 1
+	wfs.setAttrByPbEntry(&out.Attr, inode, entry)
+}
+
+func (wfs *WFS) outputFilerEntry(out *fuse.EntryOut, inode uint64, entry *filer.Entry) {
+	out.NodeId = inode
+	out.Generation = 1
+	out.EntryValid = 1
+	out.AttrValid = 1
+	wfs.setAttrByFilerEntry(&out.Attr, inode, entry)
+}
+
+func toSystemMode(mode os.FileMode) uint32 {
+	return toSystemType(mode) | uint32(mode)
+}
+
+func toSystemType(mode os.FileMode) uint32 {
+	switch mode & os.ModeType {
+	case os.ModeDir:
+		return syscall.S_IFDIR
+	case os.ModeSymlink:
+		return syscall.S_IFLNK
+	case os.ModeNamedPipe:
+		return syscall.S_IFIFO
+	case os.ModeSocket:
+		return syscall.S_IFSOCK
+	case os.ModeDevice:
+		return syscall.S_IFBLK
+	case os.ModeCharDevice:
+		return syscall.S_IFCHR
+	default:
+		return syscall.S_IFREG
+	}
+}
+
+func toFileType(mode uint32) os.FileMode {
+	switch mode & (syscall.S_IFMT & 0xffff) {
+	case syscall.S_IFDIR:
+		return os.ModeDir
+	case syscall.S_IFLNK:
+		return os.ModeSymlink
+	case syscall.S_IFIFO:
+		return os.ModeNamedPipe
+	case syscall.S_IFSOCK:
+		return os.ModeSocket
+	case syscall.S_IFBLK:
+		return os.ModeDevice
+	case syscall.S_IFCHR:
+		return os.ModeCharDevice
+	default:
+		return 0
+	}
+}
+
+func toFileMode(mode uint32) os.FileMode {
+	return toFileType(mode) | os.FileMode(mode&07777)
+}

weed/mount/weedfs_attr_darwin.go

@@ -0,0 +1,8 @@
+package mount
+
+import (
+	"github.com/hanwen/go-fuse/v2/fuse"
+)
+
+func setBlksize(out *fuse.Attr, size uint32) {
+}

weed/mount/weedfs_attr_linux.go

@@ -0,0 +1,9 @@
+package mount
+
+import (
+	"github.com/hanwen/go-fuse/v2/fuse"
+)
+
+func setBlksize(out *fuse.Attr, size uint32) {
+	out.Blksize = size
+}

weed/mount/weedfs_dir_lookup.go

@@ -0,0 +1,59 @@
+package mount
+
+import (
+	"context"
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/mount/meta_cache"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/hanwen/go-fuse/v2/fuse"
+)
+
+// Lookup is called by the kernel when the VFS wants to know
+// about a file inside a directory. Many lookup calls can
+// occur in parallel, but only one call happens for each (dir,
+// name) pair.
+
+func (wfs *WFS) Lookup(cancel <-chan struct{}, header *fuse.InHeader, name string, out *fuse.EntryOut) (code fuse.Status) {
+
+	if s := checkName(name); s != fuse.OK {
+		return s
+	}
+
+	dirPath := wfs.inodeToPath.GetPath(header.NodeId)
+
+	fullFilePath := dirPath.Child(name)
+
+	visitErr := meta_cache.EnsureVisited(wfs.metaCache, wfs, dirPath)
+	if visitErr != nil {
+		glog.Errorf("dir Lookup %s: %v", dirPath, visitErr)
+		return fuse.EIO
+	}
+	localEntry, cacheErr := wfs.metaCache.FindEntry(context.Background(), fullFilePath)
+	if cacheErr == filer_pb.ErrNotFound {
+		return fuse.ENOENT
+	}
+
+	if localEntry == nil {
+		// glog.V(3).Infof("dir Lookup cache miss %s", fullFilePath)
+		entry, err := filer_pb.GetEntry(wfs, fullFilePath)
+		if err != nil {
+			glog.V(1).Infof("dir GetEntry %s: %v", fullFilePath, err)
+			return fuse.ENOENT
+		}
+		localEntry = filer.FromPbEntry(string(dirPath), entry)
+	} else {
+		glog.V(4).Infof("dir Lookup cache hit %s", fullFilePath)
+	}
+
+	if localEntry == nil {
+		return fuse.ENOENT
+	}
+
+	inode := wfs.inodeToPath.Lookup(fullFilePath, localEntry.IsDirectory())
+
+	wfs.outputFilerEntry(out, inode, localEntry)
+
+	return fuse.OK
+
+}

weed/mount/weedfs_dir_mkrm.go

@@ -0,0 +1,111 @@
+package mount
+
+import (
+	"context"
+	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/hanwen/go-fuse/v2/fuse"
+	"os"
+	"strings"
+	"syscall"
+	"time"
+)
+
+/** Create a directory
+ *
+ * Note that the mode argument may not have the type specification
+ * bits set, i.e. S_ISDIR(mode) can be false.  To obtain the
+ * correct directory type bits use  mode|S_IFDIR
+ * */
+func (wfs *WFS) Mkdir(cancel <-chan struct{}, in *fuse.MkdirIn, name string, out *fuse.EntryOut) (code fuse.Status) {
+
+	if s := checkName(name); s != fuse.OK {
+		return s
+	}
+
+	newEntry := &filer_pb.Entry{
+		Name:        name,
+		IsDirectory: true,
+		Attributes: &filer_pb.FuseAttributes{
+			Mtime:    time.Now().Unix(),
+			Crtime:   time.Now().Unix(),
+			FileMode: uint32(os.ModeDir) | in.Mode&^uint32(wfs.option.Umask),
+			Uid:      in.Uid,
+			Gid:      in.Gid,
+		},
+	}
+
+	dirFullPath := wfs.inodeToPath.GetPath(in.NodeId)
+
+	entryFullPath := dirFullPath.Child(name)
+
+	err := wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
+
+		wfs.mapPbIdFromLocalToFiler(newEntry)
+		defer wfs.mapPbIdFromFilerToLocal(newEntry)
+
+		request := &filer_pb.CreateEntryRequest{
+			Directory:  string(dirFullPath),
+			Entry:      newEntry,
+			Signatures: []int32{wfs.signature},
+		}
+
+		glog.V(1).Infof("mkdir: %v", request)
+		if err := filer_pb.CreateEntry(client, request); err != nil {
+			glog.V(0).Infof("mkdir %s: %v", entryFullPath, err)
+			return err
+		}
+
+		if err := wfs.metaCache.InsertEntry(context.Background(), filer.FromPbEntry(request.Directory, request.Entry)); err != nil {
+			return fmt.Errorf("local mkdir dir %s: %v", entryFullPath, err)
+		}
+
+		return nil
+	})
+
+	glog.V(0).Infof("mkdir %s: %v", entryFullPath, err)
+
+	if err != nil {
+		return fuse.EIO
+	}
+
+	inode := wfs.inodeToPath.Lookup(entryFullPath, true)
+
+	wfs.outputPbEntry(out, inode, newEntry)
+
+	return fuse.OK
+
+}
+
+/** Remove a directory */
+func (wfs *WFS) Rmdir(cancel <-chan struct{}, header *fuse.InHeader, name string) (code fuse.Status) {
+
+	if name == "." {
+		return fuse.Status(syscall.EINVAL)
+	}
+	if name == ".." {
+		return fuse.Status(syscall.ENOTEMPTY)
+	}
+
+	dirFullPath := wfs.inodeToPath.GetPath(header.NodeId)
+	entryFullPath := dirFullPath.Child(name)
+
+	glog.V(3).Infof("remove directory: %v", entryFullPath)
+	ignoreRecursiveErr := true // ignore recursion error since the OS should manage it
+	err := filer_pb.Remove(wfs, string(dirFullPath), name, true, true, ignoreRecursiveErr, false, []int32{wfs.signature})
+	if err != nil {
+		glog.V(0).Infof("remove %s: %v", entryFullPath, err)
+		if strings.Contains(err.Error(), filer.MsgFailDelNonEmptyFolder) {
+			return fuse.Status(syscall.ENOTEMPTY)
+		}
+		return fuse.ENOENT
+	}
+
+	wfs.metaCache.DeleteEntry(context.Background(), entryFullPath)
+	wfs.inodeToPath.RemovePath(entryFullPath)
+
+	return fuse.OK
+
+}

weed/mount/weedfs_dir_read.go

@@ -0,0 +1,132 @@
+package mount
+
+import (
+	"context"
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/mount/meta_cache"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"github.com/hanwen/go-fuse/v2/fuse"
+	"math"
+	"os"
+)
+
+// Directory handling
+
+/** Open directory
+ *
+ * Unless the 'default_permissions' mount option is given,
+ * this method should check if opendir is permitted for this
+ * directory. Optionally opendir may also return an arbitrary
+ * filehandle in the fuse_file_info structure, which will be
+ * passed to readdir, releasedir and fsyncdir.
+ */
+func (wfs *WFS) OpenDir(cancel <-chan struct{}, input *fuse.OpenIn, out *fuse.OpenOut) (code fuse.Status) {
+	if !wfs.inodeToPath.HasInode(input.NodeId) {
+		return fuse.ENOENT
+	}
+	return fuse.OK
+}
+
+/** Release directory
+ *
+ * If the directory has been removed after the call to opendir, the
+ * path parameter will be NULL.
+ */
+func (wfs *WFS) ReleaseDir(input *fuse.ReleaseIn) {
+}
+
+/** Synchronize directory contents
+ *
+ * If the directory has been removed after the call to opendir, the
+ * path parameter will be NULL.
+ *
+ * If the datasync parameter is non-zero, then only the user data
+ * should be flushed, not the meta data
+ */
+func (wfs *WFS) FsyncDir(cancel <-chan struct{}, input *fuse.FsyncIn) (code fuse.Status) {
+	return fuse.OK
+}
+
+/** Read directory
+ *
+ * The filesystem may choose between two modes of operation:
+ *
+ * 1) The readdir implementation ignores the offset parameter, and
+ * passes zero to the filler function's offset.  The filler
+ * function will not return '1' (unless an error happens), so the
+ * whole directory is read in a single readdir operation.
+ *
+ * 2) The readdir implementation keeps track of the offsets of the
+ * directory entries.  It uses the offset parameter and always
+ * passes non-zero offset to the filler function.  When the buffer
+ * is full (or an error happens) the filler function will return
+ * '1'.
+ */
+func (wfs *WFS) ReadDir(cancel <-chan struct{}, input *fuse.ReadIn, out *fuse.DirEntryList) (code fuse.Status) {
+	return wfs.doReadDirectory(input, out, false)
+}
+
+func (wfs *WFS) ReadDirPlus(cancel <-chan struct{}, input *fuse.ReadIn, out *fuse.DirEntryList) (code fuse.Status) {
+	return wfs.doReadDirectory(input, out, true)
+}
+
+func (wfs *WFS) doReadDirectory(input *fuse.ReadIn, out *fuse.DirEntryList, isPlusMode bool) fuse.Status {
+	dirPath := wfs.inodeToPath.GetPath(input.NodeId)
+
+	var counter uint64
+	var dirEntry fuse.DirEntry
+	if input.Offset == 0 {
+		counter++
+		dirEntry.Ino = input.NodeId
+		dirEntry.Name = "."
+		dirEntry.Mode = toSystemMode(os.ModeDir)
+		out.AddDirEntry(dirEntry)
+
+		counter++
+		parentDir, _ := dirPath.DirAndName()
+		parentInode := wfs.inodeToPath.GetInode(util.FullPath(parentDir))
+		dirEntry.Ino = parentInode
+		dirEntry.Name = ".."
+		dirEntry.Mode = toSystemMode(os.ModeDir)
+		out.AddDirEntry(dirEntry)
+
+	}
+
+	processEachEntryFn := func(entry *filer.Entry, isLast bool) bool {
+		counter++
+		if counter <= input.Offset {
+			return true
+		}
+		dirEntry.Name = entry.Name()
+		inode := wfs.inodeToPath.GetInode(dirPath.Child(dirEntry.Name))
+		dirEntry.Ino = inode
+		dirEntry.Mode = toSystemMode(entry.Mode)
+		if !isPlusMode {
+			if !out.AddDirEntry(dirEntry) {
+				return false
+			}
+		} else {
+			entryOut := out.AddDirLookupEntry(dirEntry)
+			if entryOut == nil {
+				return false
+			}
+			wfs.outputFilerEntry(entryOut, inode, entry)
+		}
+		return true
+	}
+
+	if err := meta_cache.EnsureVisited(wfs.metaCache, wfs, dirPath); err != nil {
+		glog.Errorf("dir ReadDirAll %s: %v", dirPath, err)
+		return fuse.EIO
+	}
+	listErr := wfs.metaCache.ListDirectoryEntries(context.Background(), dirPath, "", false, int64(math.MaxInt32), func(entry *filer.Entry) bool {
+		return processEachEntryFn(entry, false)
+	})
+	if listErr != nil {
+		glog.Errorf("list meta cache: %v", listErr)
+		return fuse.EIO
+	}
+
+	return fuse.OK
+}

weed/mount/weedfs_file_io.go

@@ -0,0 +1,98 @@
+package mount
+
+import (
+	"github.com/hanwen/go-fuse/v2/fuse"
+)
+
+/**
+	 * Open a file
+	 *
+	 * Open flags are available in fi->flags. The following rules
+	 * apply.
+	 *
+	 *  - Creation (O_CREAT, O_EXCL, O_NOCTTY) flags will be
+	 *    filtered out / handled by the kernel.
+	 *
+	 *  - Access modes (O_RDONLY, O_WRONLY, O_RDWR) should be used
+	 *    by the filesystem to check if the operation is
+	 *    permitted.  If the ``-o default_permissions`` mount
+	 *    option is given, this check is already done by the
+	 *    kernel before calling open() and may thus be omitted by
+	 *    the filesystem.
+	 *
+	 *  - When writeback caching is enabled, the kernel may send
+	 *    read requests even for files opened with O_WRONLY. The
+	 *    filesystem should be prepared to handle this.
+	 *
+	 *  - When writeback caching is disabled, the filesystem is
+	 *    expected to properly handle the O_APPEND flag and ensure
+	 *    that each write is appending to the end of the file.
+	 *
+         *  - When writeback caching is enabled, the kernel will
+	 *    handle O_APPEND. However, unless all changes to the file
+	 *    come through the kernel this will not work reliably. The
+	 *    filesystem should thus either ignore the O_APPEND flag
+	 *    (and let the kernel handle it), or return an error
+	 *    (indicating that reliably O_APPEND is not available).
+	 *
+	 * Filesystem may store an arbitrary file handle (pointer,
+	 * index, etc) in fi->fh, and use this in other all other file
+	 * operations (read, write, flush, release, fsync).
+	 *
+	 * Filesystem may also implement stateless file I/O and not store
+	 * anything in fi->fh.
+	 *
+	 * There are also some flags (direct_io, keep_cache) which the
+	 * filesystem may set in fi, to change the way the file is opened.
+	 * See fuse_file_info structure in <fuse_common.h> for more details.
+	 *
+	 * If this request is answered with an error code of ENOSYS
+	 * and FUSE_CAP_NO_OPEN_SUPPORT is set in
+	 * `fuse_conn_info.capable`, this is treated as success and
+	 * future calls to open and release will also succeed without being
+	 * sent to the filesystem process.
+	 *
+	 * Valid replies:
+	 *   fuse_reply_open
+	 *   fuse_reply_err
+	 *
+	 * @param req request handle
+	 * @param ino the inode number
+	 * @param fi file information
+*/
+func (wfs *WFS) Open(cancel <-chan struct{}, in *fuse.OpenIn, out *fuse.OpenOut) (status fuse.Status) {
+	fileHandle, code := wfs.AcquireHandle(in.NodeId, in.Uid, in.Gid)
+	if code == fuse.OK {
+		out.Fh = uint64(fileHandle.fh)
+	}
+	return code
+}
+
+/**
+ * Release an open file
+ *
+ * Release is called when there are no more references to an open
+ * file: all file descriptors are closed and all memory mappings
+ * are unmapped.
+ *
+ * For every open call there will be exactly one release call (unless
+ * the filesystem is force-unmounted).
+ *
+ * The filesystem may reply with an error, but error values are
+ * not returned to close() or munmap() which triggered the
+ * release.
+ *
+ * fi->fh will contain the value set by the open method, or will
+ * be undefined if the open method didn't set any value.
+ * fi->flags will contain the same flags as for open.
+ *
+ * Valid replies:
+ *   fuse_reply_err
+ *
+ * @param req request handle
+ * @param ino the inode number
+ * @param fi file information
+ */
+func (wfs *WFS) Release(cancel <-chan struct{}, in *fuse.ReleaseIn) {
+	wfs.ReleaseHandle(FileHandleId(in.Fh))
+}

weed/mount/weedfs_file_mkrm.go

@@ -0,0 +1,130 @@
+package mount
+
+import (
+	"context"
+	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/hanwen/go-fuse/v2/fuse"
+	"time"
+)
+
+/**
+ * Create and open a file
+ *
+ * If the file does not exist, first create it with the specified
+ * mode, and then open it.
+ *
+ * If this method is not implemented or under Linux kernel
+ * versions earlier than 2.6.15, the mknod() and open() methods
+ * will be called instead.
+ */
+func (wfs *WFS) Create(cancel <-chan struct{}, in *fuse.CreateIn, name string, out *fuse.CreateOut) (code fuse.Status) {
+	// if implemented, need to use
+	// 	inode := wfs.inodeToPath.Lookup(entryFullPath)
+	// to ensure nlookup counter
+	return fuse.ENOSYS
+}
+
+/** Create a file node
+ *
+ * This is called for creation of all non-directory, non-symlink
+ * nodes.  If the filesystem defines a create() method, then for
+ * regular files that will be called instead.
+ */
+func (wfs *WFS) Mknod(cancel <-chan struct{}, in *fuse.MknodIn, name string, out *fuse.EntryOut) (code fuse.Status) {
+
+	if s := checkName(name); s != fuse.OK {
+		return s
+	}
+
+	newEntry := &filer_pb.Entry{
+		Name:        name,
+		IsDirectory: false,
+		Attributes: &filer_pb.FuseAttributes{
+			Mtime:       time.Now().Unix(),
+			Crtime:      time.Now().Unix(),
+			FileMode:    uint32(toFileMode(in.Mode) &^ wfs.option.Umask),
+			Uid:         in.Uid,
+			Gid:         in.Gid,
+			Collection:  wfs.option.Collection,
+			Replication: wfs.option.Replication,
+			TtlSec:      wfs.option.TtlSec,
+		},
+	}
+
+	dirFullPath := wfs.inodeToPath.GetPath(in.NodeId)
+
+	entryFullPath := dirFullPath.Child(name)
+
+	err := wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
+
+		wfs.mapPbIdFromLocalToFiler(newEntry)
+		defer wfs.mapPbIdFromFilerToLocal(newEntry)
+
+		request := &filer_pb.CreateEntryRequest{
+			Directory:  string(dirFullPath),
+			Entry:      newEntry,
+			Signatures: []int32{wfs.signature},
+		}
+
+		glog.V(1).Infof("mknod: %v", request)
+		if err := filer_pb.CreateEntry(client, request); err != nil {
+			glog.V(0).Infof("mknod %s: %v", entryFullPath, err)
+			return err
+		}
+
+		if err := wfs.metaCache.InsertEntry(context.Background(), filer.FromPbEntry(request.Directory, request.Entry)); err != nil {
+			return fmt.Errorf("local mknod %s: %v", entryFullPath, err)
+		}
+
+		return nil
+	})
+
+	glog.V(0).Infof("mknod %s: %v", entryFullPath, err)
+
+	if err != nil {
+		return fuse.EIO
+	}
+
+	inode := wfs.inodeToPath.Lookup(entryFullPath, false)
+
+	wfs.outputPbEntry(out, inode, newEntry)
+
+	return fuse.OK
+
+}
+
+/** Remove a file */
+func (wfs *WFS) Unlink(cancel <-chan struct{}, header *fuse.InHeader, name string) (code fuse.Status) {
+
+	dirFullPath := wfs.inodeToPath.GetPath(header.NodeId)
+	entryFullPath := dirFullPath.Child(name)
+
+	entry, status := wfs.maybeLoadEntry(entryFullPath)
+	if status != fuse.OK {
+		return status
+	}
+
+	// first, ensure the filer store can correctly delete
+	glog.V(3).Infof("remove file: %v", entryFullPath)
+	isDeleteData := entry != nil && entry.HardLinkCounter <= 1
+	err := filer_pb.Remove(wfs, string(dirFullPath), name, isDeleteData, false, false, false, []int32{wfs.signature})
+	if err != nil {
+		glog.V(0).Infof("remove %s: %v", entryFullPath, err)
+		return fuse.ENOENT
+	}
+
+	// then, delete meta cache
+	if err = wfs.metaCache.DeleteEntry(context.Background(), entryFullPath); err != nil {
+		glog.V(3).Infof("local DeleteEntry %s: %v", entryFullPath, err)
+		return fuse.EIO
+	}
+
+	wfs.metaCache.DeleteEntry(context.Background(), entryFullPath)
+	wfs.inodeToPath.RemovePath(entryFullPath)
+
+	return fuse.OK
+
+}

weed/mount/weedfs_file_read.go

@@ -0,0 +1,59 @@
+package mount
+
+import (
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/hanwen/go-fuse/v2/fuse"
+	"io"
+)
+
+/**
+ * Read data
+ *
+ * Read should send exactly the number of bytes requested except
+ * on EOF or error, otherwise the rest of the data will be
+ * substituted with zeroes.  An exception to this is when the file
+ * has been opened in 'direct_io' mode, in which case the return
+ * value of the read system call will reflect the return value of
+ * this operation.
+ *
+ * fi->fh will contain the value set by the open method, or will
+ * be undefined if the open method didn't set any value.
+ *
+ * Valid replies:
+ *   fuse_reply_buf
+ *   fuse_reply_iov
+ *   fuse_reply_data
+ *   fuse_reply_err
+ *
+ * @param req request handle
+ * @param ino the inode number
+ * @param size number of bytes to read
+ * @param off offset to read from
+ * @param fi file information
+ */
+func (wfs *WFS) Read(cancel <-chan struct{}, in *fuse.ReadIn, buff []byte) (fuse.ReadResult, fuse.Status) {
+	fh := wfs.GetHandle(FileHandleId(in.Fh))
+	if fh == nil {
+		return nil, fuse.ENOENT
+	}
+
+	offset := int64(in.Offset)
+	fh.lockForRead(offset, len(buff))
+	defer fh.unlockForRead(offset, len(buff))
+
+	totalRead, err := fh.readFromChunks(buff, offset)
+	if err == nil || err == io.EOF {
+		maxStop := fh.readFromDirtyPages(buff, offset)
+		totalRead = max(maxStop-offset, totalRead)
+	}
+	if err == io.EOF {
+		err = nil
+	}
+
+	if err != nil {
+		glog.Warningf("file handle read %s %d: %v", fh.FullPath(), totalRead, err)
+		return nil, fuse.EIO
+	}
+
+	return fuse.ReadResultData(buff[:totalRead]), fuse.OK
+}

weed/mount/weedfs_file_sync.go

@@ -0,0 +1,180 @@
+package mount
+
+import (
+	"context"
+	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/hanwen/go-fuse/v2/fuse"
+	"os"
+	"time"
+)
+
+/**
+ * Flush method
+ *
+ * This is called on each close() of the opened file.
+ *
+ * Since file descriptors can be duplicated (dup, dup2, fork), for
+ * one open call there may be many flush calls.
+ *
+ * Filesystems shouldn't assume that flush will always be called
+ * after some writes, or that if will be called at all.
+ *
+ * fi->fh will contain the value set by the open method, or will
+ * be undefined if the open method didn't set any value.
+ *
+ * NOTE: the name of the method is misleading, since (unlike
+ * fsync) the filesystem is not forced to flush pending writes.
+ * One reason to flush data is if the filesystem wants to return
+ * write errors during close.  However, such use is non-portable
+ * because POSIX does not require [close] to wait for delayed I/O to
+ * complete.
+ *
+ * If the filesystem supports file locking operations (setlk,
+ * getlk) it should remove all locks belonging to 'fi->owner'.
+ *
+ * If this request is answered with an error code of ENOSYS,
+ * this is treated as success and future calls to flush() will
+ * succeed automatically without being send to the filesystem
+ * process.
+ *
+ * Valid replies:
+ *   fuse_reply_err
+ *
+ * @param req request handle
+ * @param ino the inode number
+ * @param fi file information
+ *
+ * [close]: http://pubs.opengroup.org/onlinepubs/9699919799/functions/close.html
+ */
+func (wfs *WFS) Flush(cancel <-chan struct{}, in *fuse.FlushIn) fuse.Status {
+	fh := wfs.GetHandle(FileHandleId(in.Fh))
+	if fh == nil {
+		return fuse.ENOENT
+	}
+
+	fh.Lock()
+	defer fh.Unlock()
+
+	return wfs.doFlush(fh, in.Uid, in.Gid)
+}
+
+/**
+ * Synchronize file contents
+ *
+ * If the datasync parameter is non-zero, then only the user data
+ * should be flushed, not the meta data.
+ *
+ * If this request is answered with an error code of ENOSYS,
+ * this is treated as success and future calls to fsync() will
+ * succeed automatically without being send to the filesystem
+ * process.
+ *
+ * Valid replies:
+ *   fuse_reply_err
+ *
+ * @param req request handle
+ * @param ino the inode number
+ * @param datasync flag indicating if only data should be flushed
+ * @param fi file information
+ */
+func (wfs *WFS) Fsync(cancel <-chan struct{}, in *fuse.FsyncIn) (code fuse.Status) {
+
+	fh := wfs.GetHandle(FileHandleId(in.Fh))
+	if fh == nil {
+		return fuse.ENOENT
+	}
+
+	fh.Lock()
+	defer fh.Unlock()
+
+	return wfs.doFlush(fh, in.Uid, in.Gid)
+
+}
+
+func (wfs *WFS) doFlush(fh *FileHandle, uid, gid uint32) fuse.Status {
+	// flush works at fh level
+	fileFullPath := fh.FullPath()
+	dir, _ := fileFullPath.DirAndName()
+	// send the data to the OS
+	glog.V(4).Infof("doFlush %s fh %d", fileFullPath, fh.handle)
+
+	if err := fh.dirtyPages.FlushData(); err != nil {
+		glog.Errorf("%v doFlush: %v", fileFullPath, err)
+		return fuse.EIO
+	}
+
+	if !fh.dirtyMetadata {
+		return fuse.OK
+	}
+
+	err := wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
+
+		entry := fh.entry
+		if entry == nil {
+			return nil
+		}
+
+		if entry.Attributes != nil {
+			entry.Attributes.Mime = fh.contentType
+			if entry.Attributes.Uid == 0 {
+				entry.Attributes.Uid = uid
+			}
+			if entry.Attributes.Gid == 0 {
+				entry.Attributes.Gid = gid
+			}
+			if entry.Attributes.Crtime == 0 {
+				entry.Attributes.Crtime = time.Now().Unix()
+			}
+			entry.Attributes.Mtime = time.Now().Unix()
+			entry.Attributes.FileMode = uint32(os.FileMode(entry.Attributes.FileMode) &^ wfs.option.Umask)
+			entry.Attributes.Collection, entry.Attributes.Replication = fh.dirtyPages.GetStorageOptions()
+		}
+
+		request := &filer_pb.CreateEntryRequest{
+			Directory:  string(dir),
+			Entry:      entry,
+			Signatures: []int32{wfs.signature},
+		}
+
+		glog.V(4).Infof("%s set chunks: %v", fileFullPath, len(entry.Chunks))
+		for i, chunk := range entry.Chunks {
+			glog.V(4).Infof("%s chunks %d: %v [%d,%d)", fileFullPath, i, chunk.GetFileIdString(), chunk.Offset, chunk.Offset+int64(chunk.Size))
+		}
+
+		manifestChunks, nonManifestChunks := filer.SeparateManifestChunks(entry.Chunks)
+
+		chunks, _ := filer.CompactFileChunks(wfs.LookupFn(), nonManifestChunks)
+		chunks, manifestErr := filer.MaybeManifestize(wfs.saveDataAsChunk(fileFullPath), chunks)
+		if manifestErr != nil {
+			// not good, but should be ok
+			glog.V(0).Infof("MaybeManifestize: %v", manifestErr)
+		}
+		entry.Chunks = append(chunks, manifestChunks...)
+
+		wfs.mapPbIdFromLocalToFiler(request.Entry)
+		defer wfs.mapPbIdFromFilerToLocal(request.Entry)
+
+		if err := filer_pb.CreateEntry(client, request); err != nil {
+			glog.Errorf("fh flush create %s: %v", fileFullPath, err)
+			return fmt.Errorf("fh flush create %s: %v", fileFullPath, err)
+		}
+
+		wfs.metaCache.InsertEntry(context.Background(), filer.FromPbEntry(request.Directory, request.Entry))
+
+		return nil
+	})
+
+	if err == nil {
+		fh.dirtyMetadata = false
+	}
+
+	if err != nil {
+		glog.Errorf("%v fh %d flush: %v", fileFullPath, fh.handle, err)
+		return fuse.EIO
+	}
+
+	return fuse.OK
+}

weed/mount/weedfs_file_write.go

@@ -0,0 +1,66 @@
+package mount
+
+import (
+	"github.com/hanwen/go-fuse/v2/fuse"
+	"net/http"
+)
+
+/**
+ * Write data
+ *
+ * Write should return exactly the number of bytes requested
+ * except on error.  An exception to this is when the file has
+ * been opened in 'direct_io' mode, in which case the return value
+ * of the write system call will reflect the return value of this
+ * operation.
+ *
+ * Unless FUSE_CAP_HANDLE_KILLPRIV is disabled, this method is
+ * expected to reset the setuid and setgid bits.
+ *
+ * fi->fh will contain the value set by the open method, or will
+ * be undefined if the open method didn't set any value.
+ *
+ * Valid replies:
+ *   fuse_reply_write
+ *   fuse_reply_err
+ *
+ * @param req request handle
+ * @param ino the inode number
+ * @param buf data to write
+ * @param size number of bytes to write
+ * @param off offset to write to
+ * @param fi file information
+ */
+func (wfs *WFS) Write(cancel <-chan struct{}, in *fuse.WriteIn, data []byte) (written uint32, code fuse.Status) {
+
+	fh := wfs.GetHandle(FileHandleId(in.Fh))
+	if fh == nil {
+		return 0, fuse.ENOENT
+	}
+
+	fh.Lock()
+	defer fh.Unlock()
+
+	entry := fh.entry
+	if entry == nil {
+		return 0, fuse.OK
+	}
+
+	entry.Content = nil
+	offset := int64(in.Offset)
+	entry.Attributes.FileSize = uint64(max(offset+int64(len(data)), int64(entry.Attributes.FileSize)))
+	// glog.V(4).Infof("%v write [%d,%d) %d", fh.f.fullpath(), req.Offset, req.Offset+int64(len(req.Data)), len(req.Data))
+
+	fh.dirtyPages.AddPage(offset, data)
+
+	written = uint32(len(data))
+
+	if offset == 0 {
+		// detect mime type
+		fh.contentType = http.DetectContentType(data)
+	}
+
+	fh.dirtyMetadata = true
+
+	return written, fuse.OK
+}

weed/mount/weedfs_filehandle.go

@@ -0,0 +1,20 @@
+package mount
+
+import "github.com/hanwen/go-fuse/v2/fuse"
+
+func (wfs *WFS) AcquireHandle(inode uint64, uid, gid uint32) (fileHandle *FileHandle, code fuse.Status) {
+	_, _, entry, status := wfs.maybeReadEntry(inode)
+	if status == fuse.OK {
+		fileHandle = wfs.fhmap.AcquireFileHandle(wfs, inode, entry)
+		fileHandle.entry = entry
+	}
+	return
+}
+
+func (wfs *WFS) ReleaseHandle(handleId FileHandleId) {
+	wfs.fhmap.ReleaseByHandle(handleId)
+}
+
+func (wfs *WFS) GetHandle(handleId FileHandleId) *FileHandle {
+	return wfs.fhmap.GetFileHandle(handleId)
+}

weed/mount/weedfs_forget.go

@@ -0,0 +1,68 @@
+package mount
+
+import (
+	"context"
+	"github.com/chrislusf/seaweedfs/weed/util"
+)
+
+// Forget is called when the kernel discards entries from its
+// dentry cache. This happens on unmount, and when the kernel
+// is short on memory. Since it is not guaranteed to occur at
+// any moment, and since there is no return value, Forget
+// should not do I/O, as there is no channel to report back
+// I/O errors.
+// from https://github.com/libfuse/libfuse/blob/master/include/fuse_lowlevel.h
+/**
+ * Forget about an inode
+ *
+ * This function is called when the kernel removes an inode
+ * from its internal caches.
+ *
+ * The inode's lookup count increases by one for every call to
+ * fuse_reply_entry and fuse_reply_create. The nlookup parameter
+ * indicates by how much the lookup count should be decreased.
+ *
+ * Inodes with a non-zero lookup count may receive request from
+ * the kernel even after calls to unlink, rmdir or (when
+ * overwriting an existing file) rename. Filesystems must handle
+ * such requests properly and it is recommended to defer removal
+ * of the inode until the lookup count reaches zero. Calls to
+ * unlink, rmdir or rename will be followed closely by forget
+ * unless the file or directory is open, in which case the
+ * kernel issues forget only after the release or releasedir
+ * calls.
+ *
+ * Note that if a file system will be exported over NFS the
+ * inodes lifetime must extend even beyond forget. See the
+ * generation field in struct fuse_entry_param above.
+ *
+ * On unmount the lookup count for all inodes implicitly drops
+ * to zero. It is not guaranteed that the file system will
+ * receive corresponding forget messages for the affected
+ * inodes.
+ *
+ * Valid replies:
+ *   fuse_reply_none
+ *
+ * @param req request handle
+ * @param ino the inode number
+ * @param nlookup the number of lookups to forget
+ */
+/*
+https://libfuse.github.io/doxygen/include_2fuse__lowlevel_8h.html
+
+int fuse_reply_entry	(	fuse_req_t 	req,
+const struct fuse_entry_param * 	e
+)
+Reply with a directory entry
+
+Possible requests: lookup, mknod, mkdir, symlink, link
+
+Side effects: increments the lookup count on success
+
+*/
+func (wfs *WFS) Forget(nodeid, nlookup uint64) {
+	wfs.inodeToPath.Forget(nodeid, nlookup, func(dir util.FullPath) {
+		wfs.metaCache.DeleteFolderChildren(context.Background(), dir)
+	})
+}

weed/mount/weedfs_link.go

@@ -0,0 +1,93 @@
+package mount
+
+import (
+	"context"
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"github.com/hanwen/go-fuse/v2/fuse"
+	"time"
+)
+
+const (
+	HARD_LINK_MARKER = '\x01'
+)
+
+/** Create a hard link to a file */
+func (wfs *WFS) Link(cancel <-chan struct{}, in *fuse.LinkIn, name string, out *fuse.EntryOut) (code fuse.Status) {
+
+	if s := checkName(name); s != fuse.OK {
+		return s
+	}
+
+	newParentPath := wfs.inodeToPath.GetPath(in.NodeId)
+	oldEntryPath := wfs.inodeToPath.GetPath(in.Oldnodeid)
+	oldParentPath, _ := oldEntryPath.DirAndName()
+
+	oldEntry, status := wfs.maybeLoadEntry(oldEntryPath)
+	if status != fuse.OK {
+		return status
+	}
+
+	// update old file to hardlink mode
+	if len(oldEntry.HardLinkId) == 0 {
+		oldEntry.HardLinkId = append(util.RandomBytes(16), HARD_LINK_MARKER)
+		oldEntry.HardLinkCounter = 1
+	}
+	oldEntry.HardLinkCounter++
+	updateOldEntryRequest := &filer_pb.UpdateEntryRequest{
+		Directory:  oldParentPath,
+		Entry:      oldEntry,
+		Signatures: []int32{wfs.signature},
+	}
+
+	// CreateLink 1.2 : update new file to hardlink mode
+	oldEntry.Attributes.Mtime = time.Now().Unix()
+	request := &filer_pb.CreateEntryRequest{
+		Directory: string(newParentPath),
+		Entry: &filer_pb.Entry{
+			Name:            name,
+			IsDirectory:     false,
+			Attributes:      oldEntry.Attributes,
+			Chunks:          oldEntry.Chunks,
+			Extended:        oldEntry.Extended,
+			HardLinkId:      oldEntry.HardLinkId,
+			HardLinkCounter: oldEntry.HardLinkCounter,
+		},
+		Signatures: []int32{wfs.signature},
+	}
+
+	// apply changes to the filer, and also apply to local metaCache
+	err := wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
+
+		wfs.mapPbIdFromLocalToFiler(request.Entry)
+		defer wfs.mapPbIdFromFilerToLocal(request.Entry)
+
+		if err := filer_pb.UpdateEntry(client, updateOldEntryRequest); err != nil {
+			return err
+		}
+		wfs.metaCache.UpdateEntry(context.Background(), filer.FromPbEntry(updateOldEntryRequest.Directory, updateOldEntryRequest.Entry))
+
+		if err := filer_pb.CreateEntry(client, request); err != nil {
+			return err
+		}
+
+		wfs.metaCache.InsertEntry(context.Background(), filer.FromPbEntry(request.Directory, request.Entry))
+
+		return nil
+	})
+
+	newEntryPath := newParentPath.Child(name)
+
+	if err != nil {
+		glog.V(0).Infof("Link %v -> %s: %v", oldEntryPath, newEntryPath, err)
+		return fuse.EIO
+	}
+
+	inode := wfs.inodeToPath.Lookup(newEntryPath, false)
+
+	wfs.outputPbEntry(out, inode, request.Entry)
+
+	return fuse.OK
+}

weed/mount/weedfs_rename.go

@@ -0,0 +1,235 @@
+package mount
+
+import (
+	"context"
+	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"github.com/hanwen/go-fuse/v2/fs"
+	"github.com/hanwen/go-fuse/v2/fuse"
+	"io"
+	"strings"
+	"syscall"
+)
+
+/** Rename a file
+ *
+ * If the target exists it should be atomically replaced. If
+ * the target's inode's lookup count is non-zero, the file
+ * system is expected to postpone any removal of the inode
+ * until the lookup count reaches zero (see description of the
+ * forget function).
+ *
+ * If this request is answered with an error code of ENOSYS, this is
+ * treated as a permanent failure with error code EINVAL, i.e. all
+ * future bmap requests will fail with EINVAL without being
+ * send to the filesystem process.
+ *
+ * *flags* may be `RENAME_EXCHANGE` or `RENAME_NOREPLACE`. If
+ * RENAME_NOREPLACE is specified, the filesystem must not
+ * overwrite *newname* if it exists and return an error
+ * instead. If `RENAME_EXCHANGE` is specified, the filesystem
+ * must atomically exchange the two files, i.e. both must
+ * exist and neither may be deleted.
+ *
+ * Valid replies:
+ *   fuse_reply_err
+ *
+ * @param req request handle
+ * @param parent inode number of the old parent directory
+ * @param name old name
+ * @param newparent inode number of the new parent directory
+ * @param newname new name
+ */
+/*
+renameat2()
+       renameat2() has an additional flags argument.  A renameat2() call
+       with a zero flags argument is equivalent to renameat().
+
+       The flags argument is a bit mask consisting of zero or more of
+       the following flags:
+
+       RENAME_EXCHANGE
+              Atomically exchange oldpath and newpath.  Both pathnames
+              must exist but may be of different types (e.g., one could
+              be a non-empty directory and the other a symbolic link).
+
+       RENAME_NOREPLACE
+              Don't overwrite newpath of the rename.  Return an error if
+              newpath already exists.
+
+              RENAME_NOREPLACE can't be employed together with
+              RENAME_EXCHANGE.
+
+              RENAME_NOREPLACE requires support from the underlying
+              filesystem.  Support for various filesystems was added as
+              follows:
+
+              *  ext4 (Linux 3.15);
+
+              *  btrfs, tmpfs, and cifs (Linux 3.17);
+
+              *  xfs (Linux 4.0);
+
+              *  Support for many other filesystems was added in Linux
+                 4.9, including ext2, minix, reiserfs, jfs, vfat, and
+                 bpf.
+
+       RENAME_WHITEOUT (since Linux 3.18)
+              This operation makes sense only for overlay/union
+              filesystem implementations.
+
+              Specifying RENAME_WHITEOUT creates a "whiteout" object at
+              the source of the rename at the same time as performing
+              the rename.  The whole operation is atomic, so that if the
+              rename succeeds then the whiteout will also have been
+              created.
+
+              A "whiteout" is an object that has special meaning in
+              union/overlay filesystem constructs.  In these constructs,
+              multiple layers exist and only the top one is ever
+              modified.  A whiteout on an upper layer will effectively
+              hide a matching file in the lower layer, making it appear
+              as if the file didn't exist.
+
+              When a file that exists on the lower layer is renamed, the
+              file is first copied up (if not already on the upper
+              layer) and then renamed on the upper, read-write layer.
+              At the same time, the source file needs to be "whiteouted"
+              (so that the version of the source file in the lower layer
+              is rendered invisible).  The whole operation needs to be
+              done atomically.
+
+              When not part of a union/overlay, the whiteout appears as
+              a character device with a {0,0} device number.  (Note that
+              other union/overlay implementations may employ different
+              methods for storing whiteout entries; specifically, BSD
+              union mount employs a separate inode type, DT_WHT, which,
+              while supported by some filesystems available in Linux,
+              such as CODA and XFS, is ignored by the kernel's whiteout
+              support code, as of Linux 4.19, at least.)
+
+              RENAME_WHITEOUT requires the same privileges as creating a
+              device node (i.e., the CAP_MKNOD capability).
+
+              RENAME_WHITEOUT can't be employed together with
+              RENAME_EXCHANGE.
+
+              RENAME_WHITEOUT requires support from the underlying
+              filesystem.  Among the filesystems that support it are
+              tmpfs (since Linux 3.18), ext4 (since Linux 3.18), XFS
+              (since Linux 4.1), f2fs (since Linux 4.2), btrfs (since
+              Linux 4.7), and ubifs (since Linux 4.9).
+*/
+const (
+	RenameEmptyFlag = 0
+	RenameNoReplace = 1
+	RenameExchange  = fs.RENAME_EXCHANGE
+	RenameWhiteout  = 3
+)
+
+func (wfs *WFS) Rename(cancel <-chan struct{}, in *fuse.RenameIn, oldName string, newName string) (code fuse.Status) {
+	if s := checkName(newName); s != fuse.OK {
+		return s
+	}
+
+	switch in.Flags {
+	case RenameEmptyFlag:
+	case RenameNoReplace:
+	case RenameExchange:
+	case RenameWhiteout:
+		return fuse.ENOTSUP
+	default:
+		return fuse.EINVAL
+	}
+
+	oldDir := wfs.inodeToPath.GetPath(in.NodeId)
+	oldPath := oldDir.Child(oldName)
+	newDir := wfs.inodeToPath.GetPath(in.Newdir)
+	newPath := newDir.Child(newName)
+
+	glog.V(4).Infof("dir Rename %s => %s", oldPath, newPath)
+
+	// update remote filer
+	err := wfs.WithFilerClient(true, func(client filer_pb.SeaweedFilerClient) error {
+		ctx, cancel := context.WithCancel(context.Background())
+		defer cancel()
+
+		request := &filer_pb.StreamRenameEntryRequest{
+			OldDirectory: string(oldDir),
+			OldName:      oldName,
+			NewDirectory: string(newDir),
+			NewName:      newName,
+			Signatures:   []int32{wfs.signature},
+		}
+
+		stream, err := client.StreamRenameEntry(ctx, request)
+		if err != nil {
+			code = fuse.EIO
+			return fmt.Errorf("dir AtomicRenameEntry %s => %s : %v", oldPath, newPath, err)
+		}
+
+		for {
+			resp, recvErr := stream.Recv()
+			if recvErr != nil {
+				if recvErr == io.EOF {
+					break
+				} else {
+					if strings.Contains(recvErr.Error(), "not empty") {
+						code = fuse.Status(syscall.ENOTEMPTY)
+					} else if strings.Contains(recvErr.Error(), "not directory") {
+						code = fuse.ENOTDIR
+					}
+					return fmt.Errorf("dir Rename %s => %s receive: %v", oldPath, newPath, recvErr)
+				}
+			}
+
+			if err = wfs.handleRenameResponse(ctx, resp); err != nil {
+				glog.V(0).Infof("dir Rename %s => %s : %v", oldPath, newPath, err)
+				return err
+			}
+
+		}
+
+		return nil
+
+	})
+	if err != nil {
+		glog.V(0).Infof("Link: %v", err)
+		return
+	}
+
+	return fuse.OK
+
+}
+
+func (wfs *WFS) handleRenameResponse(ctx context.Context, resp *filer_pb.StreamRenameEntryResponse) error {
+	// comes from filer StreamRenameEntry, can only be create or delete entry
+
+	if resp.EventNotification.NewEntry != nil {
+		// with new entry, the old entry name also exists. This is the first step to create new entry
+		newEntry := filer.FromPbEntry(resp.EventNotification.NewParentPath, resp.EventNotification.NewEntry)
+		if err := wfs.metaCache.AtomicUpdateEntryFromFiler(ctx, "", newEntry); err != nil {
+			return err
+		}
+
+		oldParent, newParent := util.FullPath(resp.Directory), util.FullPath(resp.EventNotification.NewParentPath)
+		oldName, newName := resp.EventNotification.OldEntry.Name, resp.EventNotification.NewEntry.Name
+
+		oldPath := oldParent.Child(oldName)
+		newPath := newParent.Child(newName)
+
+		wfs.inodeToPath.MovePath(oldPath, newPath)
+
+	} else if resp.EventNotification.OldEntry != nil {
+		// without new entry, only old entry name exists. This is the second step to delete old entry
+		if err := wfs.metaCache.AtomicUpdateEntryFromFiler(ctx, util.NewFullPath(resp.Directory, resp.EventNotification.OldEntry.Name), nil); err != nil {
+			return err
+		}
+	}
+
+	return nil
+
+}

weed/mount/weedfs_stats.go

@@ -0,0 +1,80 @@
+package mount
+
+import (
+	"context"
+	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/hanwen/go-fuse/v2/fuse"
+	"math"
+	"time"
+)
+
+const blockSize = 512
+
+type statsCache struct {
+	filer_pb.StatisticsResponse
+	lastChecked int64 // unix time in seconds
+}
+
+func (wfs *WFS) StatFs(cancel <-chan struct{}, in *fuse.InHeader, out *fuse.StatfsOut) (code fuse.Status) {
+
+	glog.V(4).Infof("reading fs stats")
+
+	if wfs.stats.lastChecked < time.Now().Unix()-20 {
+
+		err := wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
+
+			request := &filer_pb.StatisticsRequest{
+				Collection:  wfs.option.Collection,
+				Replication: wfs.option.Replication,
+				Ttl:         fmt.Sprintf("%ds", wfs.option.TtlSec),
+				DiskType:    string(wfs.option.DiskType),
+			}
+
+			glog.V(4).Infof("reading filer stats: %+v", request)
+			resp, err := client.Statistics(context.Background(), request)
+			if err != nil {
+				glog.V(0).Infof("reading filer stats %v: %v", request, err)
+				return err
+			}
+			glog.V(4).Infof("read filer stats: %+v", resp)
+
+			wfs.stats.TotalSize = resp.TotalSize
+			wfs.stats.UsedSize = resp.UsedSize
+			wfs.stats.FileCount = resp.FileCount
+			wfs.stats.lastChecked = time.Now().Unix()
+
+			return nil
+		})
+		if err != nil {
+			glog.V(0).Infof("filer Statistics: %v", err)
+			return fuse.OK
+		}
+	}
+
+	totalDiskSize := wfs.stats.TotalSize
+	usedDiskSize := wfs.stats.UsedSize
+	actualFileCount := wfs.stats.FileCount
+
+	// Compute the total number of available blocks
+	out.Blocks = totalDiskSize / blockSize
+
+	// Compute the number of used blocks
+	numBlocks := uint64(usedDiskSize / blockSize)
+
+	// Report the number of free and available blocks for the block size
+	out.Bfree = out.Blocks - numBlocks
+	out.Bavail = out.Blocks - numBlocks
+	out.Bsize = uint32(blockSize)
+
+	// Report the total number of possible files in the file system (and those free)
+	out.Files = math.MaxInt64
+	out.Ffree = math.MaxInt64 - actualFileCount
+
+	// Report the maximum length of a name and the minimum fragment size
+	out.NameLen = 1024
+	out.Frsize = uint32(blockSize)
+
+	return fuse.OK
+}

weed/mount/weedfs_symlink.go

@@ -0,0 +1,78 @@
+package mount
+
+import (
+	"context"
+	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/hanwen/go-fuse/v2/fuse"
+	"os"
+	"time"
+)
+
+/** Create a symbolic link */
+func (wfs *WFS) Symlink(cancel <-chan struct{}, header *fuse.InHeader, target string, name string, out *fuse.EntryOut) (code fuse.Status) {
+
+	if s := checkName(name); s != fuse.OK {
+		return s
+	}
+
+	dirPath := wfs.inodeToPath.GetPath(header.NodeId)
+	entryFullPath := dirPath.Child(name)
+
+	request := &filer_pb.CreateEntryRequest{
+		Directory: string(dirPath),
+		Entry: &filer_pb.Entry{
+			Name:        name,
+			IsDirectory: false,
+			Attributes: &filer_pb.FuseAttributes{
+				Mtime:         time.Now().Unix(),
+				Crtime:        time.Now().Unix(),
+				FileMode:      uint32((os.FileMode(0777) | os.ModeSymlink) &^ wfs.option.Umask),
+				Uid:           header.Uid,
+				Gid:           header.Gid,
+				SymlinkTarget: target,
+			},
+		},
+		Signatures: []int32{wfs.signature},
+	}
+
+	err := wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
+
+		wfs.mapPbIdFromLocalToFiler(request.Entry)
+		defer wfs.mapPbIdFromFilerToLocal(request.Entry)
+
+		if err := filer_pb.CreateEntry(client, request); err != nil {
+			return fmt.Errorf("symlink %s: %v", entryFullPath, err)
+		}
+
+		wfs.metaCache.InsertEntry(context.Background(), filer.FromPbEntry(request.Directory, request.Entry))
+
+		return nil
+	})
+	if err != nil {
+		glog.V(0).Infof("Symlink %s => %s: %v", entryFullPath, target, err)
+		return fuse.EIO
+	}
+
+	inode := wfs.inodeToPath.Lookup(entryFullPath, false)
+
+	wfs.outputPbEntry(out, inode, request.Entry)
+
+	return fuse.OK
+}
+
+func (wfs *WFS) Readlink(cancel <-chan struct{}, header *fuse.InHeader) (out []byte, code fuse.Status) {
+	entryFullPath := wfs.inodeToPath.GetPath(header.NodeId)
+
+	entry, status := wfs.maybeLoadEntry(entryFullPath)
+	if status != fuse.OK {
+		return nil, status
+	}
+	if os.FileMode(entry.Attributes.FileMode)&os.ModeSymlink == 0 {
+		return nil, fuse.EINVAL
+	}
+
+	return []byte(entry.Attributes.SymlinkTarget), fuse.OK
+}

weed/mount/weedfs_unsupported.go

@@ -0,0 +1,65 @@
+package mount
+
+import "github.com/hanwen/go-fuse/v2/fuse"
+
+// https://github.com/libfuse/libfuse/blob/48ae2e72b39b6a31cb2194f6f11786b7ca06aac6/include/fuse.h#L778
+
+/**
+ * Copy a range of data from one file to anotherNiels de Vos, 4 years ago: • libfuse: add copy_file_range() support
+ *
+ * Performs an optimized copy between two file descriptors without the
+ * additional cost of transferring data through the FUSE kernel module
+ * to user space (glibc) and then back into the FUSE filesystem again.
+ *
+ * In case this method is not implemented, applications are expected to
+ * fall back to a regular file copy.   (Some glibc versions did this
+ * emulation automatically, but the emulation has been removed from all
+ * glibc release branches.)
+ */
+func (wfs *WFS) CopyFileRange(cancel <-chan struct{}, in *fuse.CopyFileRangeIn) (written uint32, code fuse.Status) {
+	return 0, fuse.ENOSYS
+}
+
+/**
+ * Allocates space for an open file
+ *
+ * This function ensures that required space is allocated for specified
+ * file.  If this function returns success then any subsequent write
+ * request to specified range is guaranteed not to fail because of lack
+ * of space on the file system media.
+ */
+func (wfs *WFS) Fallocate(cancel <-chan struct{}, in *fuse.FallocateIn) (code fuse.Status) {
+	return fuse.ENOSYS
+}
+
+/**
+ * Find next data or hole after the specified offset
+ */
+func (wfs *WFS) Lseek(cancel <-chan struct{}, in *fuse.LseekIn, out *fuse.LseekOut) fuse.Status {
+	return fuse.ENOSYS
+}
+
+func (wfs *WFS) GetLk(cancel <-chan struct{}, in *fuse.LkIn, out *fuse.LkOut) (code fuse.Status) {
+	return fuse.ENOSYS
+}
+
+func (wfs *WFS) SetLk(cancel <-chan struct{}, in *fuse.LkIn) (code fuse.Status) {
+	return fuse.ENOSYS
+}
+
+func (wfs *WFS) SetLkw(cancel <-chan struct{}, in *fuse.LkIn) (code fuse.Status) {
+	return fuse.ENOSYS
+}
+
+/**
+ * Check file access permissions
+ *
+ * This will be called for the access() system call.  If the
+ * 'default_permissions' mount option is given, this method is not
+ * called.
+ *
+ * This method is not called under Linux kernel versions 2.4.x
+ */
+func (wfs *WFS) Access(cancel <-chan struct{}, input *fuse.AccessIn) (code fuse.Status) {
+	return fuse.ENOSYS
+}

weed/mount/weedfs_write.go

@@ -0,0 +1,84 @@
+package mount
+
+import (
+	"context"
+	"fmt"
+	"io"
+
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/operation"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/chrislusf/seaweedfs/weed/security"
+	"github.com/chrislusf/seaweedfs/weed/util"
+)
+
+func (wfs *WFS) saveDataAsChunk(fullPath util.FullPath) filer.SaveDataAsChunkFunctionType {
+
+	return func(reader io.Reader, filename string, offset int64) (chunk *filer_pb.FileChunk, collection, replication string, err error) {
+		var fileId, host string
+		var auth security.EncodedJwt
+
+		if err := wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
+			return util.Retry("assignVolume", func() error {
+				request := &filer_pb.AssignVolumeRequest{
+					Count:       1,
+					Replication: wfs.option.Replication,
+					Collection:  wfs.option.Collection,
+					TtlSec:      wfs.option.TtlSec,
+					DiskType:    string(wfs.option.DiskType),
+					DataCenter:  wfs.option.DataCenter,
+					Path:        string(fullPath),
+				}
+
+				resp, err := client.AssignVolume(context.Background(), request)
+				if err != nil {
+					glog.V(0).Infof("assign volume failure %v: %v", request, err)
+					return err
+				}
+				if resp.Error != "" {
+					return fmt.Errorf("assign volume failure %v: %v", request, resp.Error)
+				}
+
+				fileId, auth = resp.FileId, security.EncodedJwt(resp.Auth)
+				loc := resp.Location
+				host = wfs.AdjustedUrl(loc)
+				collection, replication = resp.Collection, resp.Replication
+
+				return nil
+			})
+		}); err != nil {
+			return nil, "", "", fmt.Errorf("filerGrpcAddress assign volume: %v", err)
+		}
+
+		fileUrl := fmt.Sprintf("http://%s/%s", host, fileId)
+		if wfs.option.VolumeServerAccess == "filerProxy" {
+			fileUrl = fmt.Sprintf("http://%s/?proxyChunkId=%s", wfs.getCurrentFiler(), fileId)
+		}
+		uploadOption := &operation.UploadOption{
+			UploadUrl:         fileUrl,
+			Filename:          filename,
+			Cipher:            wfs.option.Cipher,
+			IsInputCompressed: false,
+			MimeType:          "",
+			PairMap:           nil,
+			Jwt:               auth,
+		}
+		uploadResult, err, data := operation.Upload(reader, uploadOption)
+		if err != nil {
+			glog.V(0).Infof("upload data %v to %s: %v", filename, fileUrl, err)
+			return nil, "", "", fmt.Errorf("upload data: %v", err)
+		}
+		if uploadResult.Error != "" {
+			glog.V(0).Infof("upload failure %v to %s: %v", filename, fileUrl, err)
+			return nil, "", "", fmt.Errorf("upload result: %v", uploadResult.Error)
+		}
+
+		if offset == 0 {
+			wfs.chunkCache.SetChunk(fileId, data)
+		}
+
+		chunk = uploadResult.ToPbFileChunk(fileId, offset)
+		return chunk, collection, replication, nil
+	}
+}

weed/mount/weedfs_xattr.go

@@ -0,0 +1,168 @@
+package mount
+
+import (
+	"github.com/hanwen/go-fuse/v2/fuse"
+	sys "golang.org/x/sys/unix"
+	"runtime"
+	"strings"
+	"syscall"
+)
+
+const (
+	// https://man7.org/linux/man-pages/man7/xattr.7.html#:~:text=The%20VFS%20imposes%20limitations%20that,in%20listxattr(2)).
+	MAX_XATTR_NAME_SIZE  = 255
+	MAX_XATTR_VALUE_SIZE = 65536
+	XATTR_PREFIX         = "xattr-" // same as filer
+)
+
+// GetXAttr reads an extended attribute, and should return the
+// number of bytes. If the buffer is too small, return ERANGE,
+// with the required buffer size.
+func (wfs *WFS) GetXAttr(cancel <-chan struct{}, header *fuse.InHeader, attr string, dest []byte) (size uint32, code fuse.Status) {
+
+	//validate attr name
+	if len(attr) > MAX_XATTR_NAME_SIZE {
+		if runtime.GOOS == "darwin" {
+			return 0, fuse.EPERM
+		} else {
+			return 0, fuse.ERANGE
+		}
+	}
+	if len(attr) == 0 {
+		return 0, fuse.EINVAL
+	}
+
+	_, _, entry, status := wfs.maybeReadEntry(header.NodeId)
+	if status != fuse.OK {
+		return 0, status
+	}
+	if entry == nil {
+		return 0, fuse.ENOENT
+	}
+	if entry.Extended == nil {
+		return 0, fuse.ENOATTR
+	}
+	data, found := entry.Extended[XATTR_PREFIX+attr]
+	if !found {
+		return 0, fuse.ENOATTR
+	}
+	if len(dest) < len(data) {
+		return uint32(len(data)), fuse.ERANGE
+	}
+	copy(dest, data)
+
+	return uint32(len(data)), fuse.OK
+}
+
+// SetXAttr writes an extended attribute.
+// https://man7.org/linux/man-pages/man2/setxattr.2.html
+//        By default (i.e., flags is zero), the extended attribute will be
+//       created if it does not exist, or the value will be replaced if
+//       the attribute already exists.  To modify these semantics, one of
+//       the following values can be specified in flags:
+//
+//       XATTR_CREATE
+//              Perform a pure create, which fails if the named attribute
+//              exists already.
+//
+//       XATTR_REPLACE
+//              Perform a pure replace operation, which fails if the named
+//              attribute does not already exist.
+func (wfs *WFS) SetXAttr(cancel <-chan struct{}, input *fuse.SetXAttrIn, attr string, data []byte) fuse.Status {
+	//validate attr name
+	if len(attr) > MAX_XATTR_NAME_SIZE {
+		if runtime.GOOS == "darwin" {
+			return fuse.EPERM
+		} else {
+			return fuse.ERANGE
+		}
+	}
+	if len(attr) == 0 {
+		return fuse.EINVAL
+	}
+	//validate attr value
+	if len(data) > MAX_XATTR_VALUE_SIZE {
+		if runtime.GOOS == "darwin" {
+			return fuse.Status(syscall.E2BIG)
+		} else {
+			return fuse.ERANGE
+		}
+	}
+
+	path, _, entry, status := wfs.maybeReadEntry(input.NodeId)
+	if status != fuse.OK {
+		return status
+	}
+	if entry.Extended == nil {
+		entry.Extended = make(map[string][]byte)
+	}
+	oldData, _ := entry.Extended[XATTR_PREFIX+attr]
+	switch input.Flags {
+	case sys.XATTR_CREATE:
+		if len(oldData) > 0 {
+			break
+		}
+		fallthrough
+	case sys.XATTR_REPLACE:
+		fallthrough
+	default:
+		entry.Extended[XATTR_PREFIX+attr] = data
+	}
+
+	return wfs.saveEntry(path, entry)
+
+}
+
+// ListXAttr lists extended attributes as '\0' delimited byte
+// slice, and return the number of bytes. If the buffer is too
+// small, return ERANGE, with the required buffer size.
+func (wfs *WFS) ListXAttr(cancel <-chan struct{}, header *fuse.InHeader, dest []byte) (n uint32, code fuse.Status) {
+	_, _, entry, status := wfs.maybeReadEntry(header.NodeId)
+	if status != fuse.OK {
+		return 0, status
+	}
+	if entry == nil {
+		return 0, fuse.ENOENT
+	}
+	if entry.Extended == nil {
+		return 0, fuse.ENOATTR
+	}
+
+	var data []byte
+	for k := range entry.Extended {
+		if strings.HasPrefix(k, XATTR_PREFIX) {
+			data = append(data, k[len(XATTR_PREFIX):]...)
+			data = append(data, 0)
+		}
+	}
+	if len(dest) < len(data) {
+		return uint32(len(data)), fuse.ERANGE
+	}
+
+	copy(dest, data)
+
+	return uint32(len(data)), fuse.OK
+}
+
+// RemoveXAttr removes an extended attribute.
+func (wfs *WFS) RemoveXAttr(cancel <-chan struct{}, header *fuse.InHeader, attr string) fuse.Status {
+	if len(attr) == 0 {
+		return fuse.EINVAL
+	}
+	path, _, entry, status := wfs.maybeReadEntry(header.NodeId)
+	if status != fuse.OK {
+		return status
+	}
+	if entry.Extended == nil {
+		return fuse.ENOATTR
+	}
+	_, found := entry.Extended[XATTR_PREFIX+attr]
+
+	if !found {
+		return fuse.ENOATTR
+	}
+
+	delete(entry.Extended, XATTR_PREFIX+attr)
+
+	return wfs.saveEntry(path, entry)
+}

weed/mount/wfs_filer_client.go

@@ -0,0 +1,51 @@
+package mount
+
+import (
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"google.golang.org/grpc"
+
+	"github.com/chrislusf/seaweedfs/weed/pb"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+)
+
+var _ = filer_pb.FilerClient(&WFS{})
+
+func (wfs *WFS) WithFilerClient(streamingMode bool, fn func(filer_pb.SeaweedFilerClient) error) (err error) {
+
+	return util.Retry("filer grpc", func() error {
+
+		i := wfs.option.filerIndex
+		n := len(wfs.option.FilerAddresses)
+		for x := 0; x < n; x++ {
+
+			filerGrpcAddress := wfs.option.FilerAddresses[i].ToGrpcAddress()
+			err = pb.WithGrpcClient(streamingMode, func(grpcConnection *grpc.ClientConn) error {
+				client := filer_pb.NewSeaweedFilerClient(grpcConnection)
+				return fn(client)
+			}, filerGrpcAddress, wfs.option.GrpcDialOption)
+
+			if err != nil {
+				glog.V(0).Infof("WithFilerClient %d %v: %v", x, filerGrpcAddress, err)
+			} else {
+				wfs.option.filerIndex = i
+				return nil
+			}
+
+			i++
+			if i >= n {
+				i = 0
+			}
+
+		}
+		return err
+	})
+
+}
+
+func (wfs *WFS) AdjustedUrl(location *filer_pb.Location) string {
+	if wfs.option.VolumeServerAccess == "publicUrl" {
+		return location.PublicUrl
+	}
+	return location.Url
+}

weed/mount/wfs_save.go

@@ -0,0 +1,67 @@
+package mount
+
+import (
+	"context"
+	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"github.com/hanwen/go-fuse/v2/fuse"
+	"syscall"
+)
+
+func (wfs *WFS) saveEntry(path util.FullPath, entry *filer_pb.Entry) (code fuse.Status) {
+
+	parentDir, _ := path.DirAndName()
+
+	err := wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
+
+		wfs.mapPbIdFromLocalToFiler(entry)
+		defer wfs.mapPbIdFromFilerToLocal(entry)
+
+		request := &filer_pb.UpdateEntryRequest{
+			Directory:  parentDir,
+			Entry:      entry,
+			Signatures: []int32{wfs.signature},
+		}
+
+		glog.V(1).Infof("save entry: %v", request)
+		_, err := client.UpdateEntry(context.Background(), request)
+		if err != nil {
+			return fmt.Errorf("UpdateEntry dir %s: %v", path, err)
+		}
+
+		if err := wfs.metaCache.UpdateEntry(context.Background(), filer.FromPbEntry(request.Directory, request.Entry)); err != nil {
+			return fmt.Errorf("UpdateEntry dir %s: %v", path, err)
+		}
+
+		return nil
+	})
+	if err != nil {
+		glog.Errorf("saveEntry %s: %v", path, err)
+		return fuse.EIO
+	}
+
+	return fuse.OK
+}
+
+func (wfs *WFS) mapPbIdFromFilerToLocal(entry *filer_pb.Entry) {
+	if entry.Attributes == nil {
+		return
+	}
+	entry.Attributes.Uid, entry.Attributes.Gid = wfs.option.UidGidMapper.FilerToLocal(entry.Attributes.Uid, entry.Attributes.Gid)
+}
+func (wfs *WFS) mapPbIdFromLocalToFiler(entry *filer_pb.Entry) {
+	if entry.Attributes == nil {
+		return
+	}
+	entry.Attributes.Uid, entry.Attributes.Gid = wfs.option.UidGidMapper.LocalToFiler(entry.Attributes.Uid, entry.Attributes.Gid)
+}
+
+func checkName(name string) fuse.Status {
+	if len(name) >= 256 {
+		return fuse.Status(syscall.ENAMETOOLONG)
+	}
+	return fuse.OK
+}

weed/pb/filer.proto

@@ -114,6 +114,7 @@ message Entry {
     bytes content = 9; // if not empty, the file content
 
     RemoteEntry remote_entry = 10;
+    int64 quota = 11; // for bucket only. Positive/Negative means enabled/disabled.
 }
 
 message FullEntry {

weed/pb/grpc_client_server.go

@@ -45,9 +45,8 @@ func NewGrpcServer(opts ...grpc.ServerOption) *grpc.Server {
 	var options []grpc.ServerOption
 	options = append(options,
 		grpc.KeepaliveParams(keepalive.ServerParameters{
-			Time:             10 * time.Second, // wait time before ping if no activity
-			Timeout:          20 * time.Second, // ping timeout
-			MaxConnectionAge: 10 * time.Hour,
+			Time:    10 * time.Second, // wait time before ping if no activity
+			Timeout: 20 * time.Second, // ping timeout
 		}),
 		grpc.KeepaliveEnforcementPolicy(keepalive.EnforcementPolicy{
 			MinTime:             60 * time.Second, // min time a client should wait before sending a ping

weed/replication/sink/filersink/filer_sink.go

@@ -199,7 +199,7 @@ func (fs *FilerSink) UpdateEntry(key string, oldEntry *filer_pb.Entry, newParent
 		// delete the chunks that are deleted from the source
 		if deleteIncludeChunks {
 			// remove the deleted chunks. Actual data deletion happens in filer UpdateEntry FindUnusedFileChunks
-			existingEntry.Chunks = filer.DoMinusChunks(existingEntry.Chunks, deletedChunks)
+			existingEntry.Chunks = filer.DoMinusChunksBySourceFileId(existingEntry.Chunks, deletedChunks)
 		}
 
 		// replicate the chunks that are new in the source

weed/replication/sub/notification_kafka.go

@@ -137,7 +137,7 @@ func (progress *KafkaProgress) saveProgress() error {
 	if err != nil {
 		return fmt.Errorf("failed to marshal progress: %v", err)
 	}
-	err = os.WriteFile(progress.offsetFile, data, 0640)
+	err = util.WriteFile(progress.offsetFile, data, 0640)
 	if err != nil {
 		return fmt.Errorf("failed to save progress to %s: %v", progress.offsetFile, err)
 	}

weed/s3api/s3api_bucket_handlers.go

@@ -4,13 +4,14 @@ import (
 	"context"
 	"encoding/xml"
 	"fmt"
-	"github.com/chrislusf/seaweedfs/weed/filer"
-	"github.com/chrislusf/seaweedfs/weed/s3api/s3_constants"
-	"github.com/chrislusf/seaweedfs/weed/storage/needle"
 	"math"
 	"net/http"
 	"time"
 
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/s3api/s3_constants"
+	"github.com/chrislusf/seaweedfs/weed/storage/needle"
+
 	xhttp "github.com/chrislusf/seaweedfs/weed/s3api/http"
 	"github.com/chrislusf/seaweedfs/weed/s3api/s3err"
 
@@ -309,3 +310,15 @@ func (s3a *S3ApiServer) DeleteBucketLifecycleHandler(w http.ResponseWriter, r *h
 	s3err.WriteEmptyResponse(w, r, http.StatusNoContent)
 
 }
+
+// GetBucketLocationHandler Get bucket location
+// https://docs.aws.amazon.com/AmazonS3/latest/API/API_GetBucketLocation.html
+func (s3a *S3ApiServer) GetBucketLocationHandler(w http.ResponseWriter, r *http.Request) {
+	writeSuccessResponseXML(w, r, LocationConstraint{})
+}
+
+// GetBucketRequestPaymentHandler Get bucket location
+// https://docs.aws.amazon.com/AmazonS3/latest/API/API_GetBucketRequestPayment.html
+func (s3a *S3ApiServer) GetBucketRequestPaymentHandler(w http.ResponseWriter, r *http.Request) {
+	writeSuccessResponseXML(w, r, RequestPaymentConfiguration{Payer: "BucketOwner"})
+}