Zach Brown e14912974d Wait for lock recovery before sending farewell
We recently found that the server can send a farewell response and try
to tear down a client's lock state while it was still in lock recovery
with the client.   The lock recovery response could add a lock
for the client after farell's reclaim_rid() had thought the client was
gone forever and tore down its locks.

This left a lock in the lock server that wasn't associated with any
clients and so could never be invalidated.   Attempts to acquire
conflicting locks with it would hang forever, which we saw as hangs in
testing with lots of unmounting.

We tried to fix it by serializing incoming request handling and
forcefully clobbering the client's lock state as we first got
the farewell request.   That went very badly.

This takes another approach of trying to explicitly wait for lock
recovery to finish before sending farewell responses.   It's more in
line with the overall pattern of having the client be up and functional
until farewell tears it down.

With this in place we can revert the other attempted fix that was
causing so many problems.

Signed-off-by: Zach Brown <zab@versity.com>
2022-01-24 09:39:51 -08:00
2020-12-07 09:47:12 -08:00
2020-12-07 10:39:20 -08:00
2021-11-05 11:16:57 -07:00

Introduction

scoutfs is a clustered in-kernel Linux filesystem designed to support large archival systems. It features additional interfaces and metadata so that archive agents can perform their maintenance workflows without walking all the files in the namespace. Its cluster support lets deployments add nodes to satisfy archival tier bandwidth targets.

The design goal is to reach file populations in the trillions, with the archival bandwidth to match, while remaining operational and responsive.

Highlights of the design and implementation include:

  • Fully consistent POSIX semantics between nodes
  • Atomic transactions to maintain consistent persistent structures
  • Integrated archival metadata replaces syncing to external databases
  • Dynamic seperation of resources lets nodes write in parallel
  • 64bit throughout; no limits on file or directory sizes or counts
  • Open GPLv2 implementation

Community Mailing List

Please join us on the open scoutfs-devel@scoutfs.org mailing list hosted on Google Groups

Description
No description provided
Readme 6.6 MiB
Languages
C 86.4%
Shell 10%
Roff 2.5%
TeX 0.8%
Makefile 0.3%