Files
seaweedfs/weed/server/master_server_handlers_admin.go
T
Chris LuGitHubDevin <158243242+devin-ai-integration[bot]@users.noreply.github.com>Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
a261f90e18 vacuum: let the sweep release volumes that stay empty and quiet (#11477)
* vacuum: let the sweep release volumes that stay empty and quiet

Vacuuming reclaims bytes but not slots: a fully emptied volume stays
registered to its collection forever, and since growth is gated only on
slot count a store at 99% free disk can still refuse writes to other
collections (#11429). volume.deleteEmpty exists but is manual-only.

With -vacuumDeleteEmptyAfterSeconds (or master.vacuumDeleteEmptyAfterSeconds
under weed server/mini; default 0, off) the automatic sweep now deletes
replica copies that have stayed empty and quiet for that long, the same
rule volume.deleteEmpty applies on demand: remote-backed copies are
skipped, and every delete carries the volume server's onlyEmpty /
onlyGarbage guards so a copy written since the last report is refused
rather than removed. Copies that still hold data or were written
recently stay; only a volume whose every copy is deleted leaves the
sweep's work map, sparing a compaction of bytes that are all deleted.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* vacuum: harden empty-volume sweep against partial and racing deletes

Review follow-up on #11477:

- delete a volume only when every replica copy is a verifiable
  empty-and-quiet candidate; deleting the empty copy of a volume whose
  sibling holds live files would silently cut its replica count
  (greptile P1).
- drain the volume out of the writable list before deleting, the same
  drain the compact pass uses, so PickForWrite stops assigning it and
  pending writes settle (devin).
- bound the VolumeDelete RPC so one stalled server cannot hold the
  vacuum lock indefinitely (greptile P1, reusing allocateVolumeTimeout).

The vid2location panic scenario raised in review does not exist:
VolumeLocationList methods are nil-receiver safe and a missing vid just
fails enoughCopies, so a partially deleted volume skips compaction
instead of crashing the sweep.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* vacuum: unregister deleted empty replicas and prune the sweep list

A successful VolumeDelete only updates the volume server; the master
still tracked the replica and kept it in the sweep's location list for
the compaction pass (coderabbit on #11477). Unregister the replica right
after its delete succeeds and drop it from the sweep copy, so a partially
deleted volume only compacts copies that still exist.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* vacuum: pin deleting volumes out of the writable list across heartbeats

Review follow-up on #11477 (greptile): DrainAndRemoveFromWritable only
removed the volume once; a heartbeat landing between the drain and the
replica deletes re-evaluated writability and re-added it, so a client
write could reach a replica whose siblings were already gone and leave
the volume under-replicated when the last copy refused its onlyEmpty
delete.

MarkDeleting records the vid in deletingVolumes — checked inside
setVolumeWritable so heartbeat, capacity-recovery, and admin re-add
paths all hold it out — and UnmarkDeleting releases it once the sweep
finishes the copy pass. A partially deleted volume's surviving replicas
then return to writable through the normal heartbeat path.

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

* vacuum: restore writability when a sweep delete survives

Generated with [Devin](https://devin.ai)

Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>

---------

Co-authored-by: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
2026-09-27 07:05:22 +08:00

237 lines
8.4 KiB
Go

package weed_server
import (
"context"
"fmt"
"math/rand/v2"
"net/http"
"strconv"
"github.com/seaweedfs/seaweedfs/weed/util/version"
"github.com/seaweedfs/seaweedfs/weed/pb"
"github.com/seaweedfs/seaweedfs/weed/pb/master_pb"
"github.com/seaweedfs/seaweedfs/weed/glog"
"github.com/seaweedfs/seaweedfs/weed/storage/backend/memory_map"
"github.com/seaweedfs/seaweedfs/weed/storage/needle"
"github.com/seaweedfs/seaweedfs/weed/storage/super_block"
"github.com/seaweedfs/seaweedfs/weed/storage/types"
"github.com/seaweedfs/seaweedfs/weed/topology"
util_http "github.com/seaweedfs/seaweedfs/weed/util/http"
)
func (ms *MasterServer) collectionDeleteHandler(w http.ResponseWriter, r *http.Request) {
collectionName := r.FormValue("collection")
if _, ok := ms.Topo.FindCollection(collectionName); !ok {
writeJsonError(w, r, http.StatusBadRequest, fmt.Errorf("collection %s does not exist", collectionName))
return
}
// The same two passes the gRPC CollectionDelete runs, rather than a second
// copy of the volume-server walk: this handler deleted only the normal
// volumes and left the collection's EC shards behind.
if err := ms.deleteCollection(r.Context(), collectionName); err != nil {
writeJsonError(w, r, http.StatusInternalServerError, err)
return
}
w.WriteHeader(http.StatusNoContent)
return
}
func (ms *MasterServer) dirStatusHandler(w http.ResponseWriter, r *http.Request) {
m := make(map[string]interface{})
m["Version"] = version.Version()
m["TopologyId"] = ms.Topo.GetTopologyId()
m["Topology"] = ms.Topo.ToInfo()
writeJsonQuiet(w, r, http.StatusOK, m)
}
func (ms *MasterServer) volumeVacuumHandler(w http.ResponseWriter, r *http.Request) {
gcString := r.FormValue("garbageThreshold")
gcThreshold := ms.option.GarbageThreshold
if gcString != "" {
var err error
gcThreshold, err = strconv.ParseFloat(gcString, 32)
if err != nil {
glog.V(0).Infof("garbageThreshold %s is not a valid float number: %v", gcString, err)
writeJsonError(w, r, http.StatusNotAcceptable, fmt.Errorf("garbageThreshold %s is not a valid float number", gcString))
return
}
}
// glog.Infoln("garbageThreshold =", gcThreshold)
ms.Topo.Vacuum(ms.grpcDialOption, gcThreshold, ms.option.MaxParallelVacuumPerServer, 0, "", ms.preallocateSize, false, 0)
ms.dirStatusHandler(w, r)
}
func (ms *MasterServer) volumeGrowHandler(w http.ResponseWriter, r *http.Request) {
count := uint64(0)
option, err := ms.getVolumeGrowOption(r)
if err != nil {
writeJsonError(w, r, http.StatusNotAcceptable, err)
return
}
glog.V(0).Infof("volumeGrowHandler received %v from %v", option.String(), r.RemoteAddr)
if count, err = strconv.ParseUint(r.FormValue("count"), 10, 32); err == nil {
replicaCount := int64(count * uint64(option.ReplicaPlacement.GetCopyCount()))
if ms.Topo.AvailableSpaceFor(option) < replicaCount {
err = fmt.Errorf("only %d volumes left, not enough for %d", ms.Topo.AvailableSpaceFor(option), replicaCount)
} else if !ms.Topo.DataCenterExists(option.DataCenter) {
err = fmt.Errorf("data center %v not found in topology", option.DataCenter)
} else {
var newVidLocations []*master_pb.VolumeLocation
newVidLocations, err = ms.vg.GrowByCountAndType(ms.grpcDialOption, uint32(count), option, ms.Topo)
if len(newVidLocations) > 0 {
ms.broadcastVolumeLocationsToClients(newVidLocations)
}
count = uint64(len(newVidLocations))
}
} else {
err = fmt.Errorf("can not parse parameter count %s", r.FormValue("count"))
}
if err != nil {
writeJsonError(w, r, http.StatusNotAcceptable, err)
} else {
writeJsonQuiet(w, r, http.StatusOK, map[string]interface{}{"count": count})
}
}
func (ms *MasterServer) volumeStatusHandler(w http.ResponseWriter, r *http.Request) {
m := make(map[string]interface{})
m["Version"] = version.Version()
m["Volumes"] = ms.Topo.ToVolumeMap()
writeJsonQuiet(w, r, http.StatusOK, m)
}
func (ms *MasterServer) redirectHandler(w http.ResponseWriter, r *http.Request) {
vid, _, _, _, _ := parseURLPath(r.URL.Path)
collection := r.FormValue("collection")
location := ms.findVolumeLocation(collection, vid)
if location.Error == "" {
loc := location.Locations[rand.IntN(len(location.Locations))]
url, _ := util_http.NormalizeUrl(loc.PublicUrl)
if r.URL.RawQuery != "" {
url = url + r.URL.Path + "?" + r.URL.RawQuery
} else {
url = url + r.URL.Path
}
http.Redirect(w, r, url, http.StatusPermanentRedirect)
} else {
writeJsonError(w, r, http.StatusNotFound, fmt.Errorf("volume id %s not found: %s", vid, location.Error))
}
}
func (ms *MasterServer) submitFromMasterServerHandler(w http.ResponseWriter, r *http.Request) {
fileSizeLimitBytes := int64(ms.option.FileSizeLimitMB) * 1024 * 1024
if ms.Topo.IsLeader() {
submitForClientHandler(w, r, func(ctx context.Context) pb.ServerAddress { return ms.option.Master }, ms.grpcDialOption, fileSizeLimitBytes)
} else {
masterUrl, err := ms.Topo.Leader()
if err != nil {
writeJsonError(w, r, http.StatusInternalServerError, err)
} else {
submitForClientHandler(w, r, func(ctx context.Context) pb.ServerAddress { return masterUrl }, ms.grpcDialOption, fileSizeLimitBytes)
}
}
}
func (ms *MasterServer) getVolumeGrowOption(r *http.Request) (*topology.VolumeGrowOption, error) {
replicationString := r.FormValue("replication")
if replicationString == "" {
replicationString = ms.option.DefaultReplicaPlacement
}
replicaPlacement, err := super_block.NewReplicaPlacementFromString(replicationString)
if err != nil {
return nil, err
}
ttl, err := needle.ReadTTL(r.FormValue("ttl"))
if err != nil {
return nil, err
}
memoryMapMaxSizeMb, err := memory_map.ReadMemoryMapMaxSizeMb(r.FormValue("memoryMapMaxSizeMb"))
if err != nil {
return nil, err
}
diskType := types.ToDiskType(r.FormValue("disk"))
preallocate := ms.preallocateSize
if r.FormValue("preallocate") != "" {
preallocate, err = strconv.ParseInt(r.FormValue("preallocate"), 10, 64)
if err != nil {
return nil, fmt.Errorf("Failed to parse int64 preallocate = %s: %v", r.FormValue("preallocate"), err)
}
}
ver := needle.GetCurrentVersion()
volumeGrowOption := &topology.VolumeGrowOption{
Collection: r.FormValue("collection"),
ReplicaPlacement: replicaPlacement,
Ttl: ttl,
DiskType: diskType,
Preallocate: preallocate,
DataCenter: r.FormValue("dataCenter"),
Rack: r.FormValue("rack"),
DataNode: r.FormValue("dataNode"),
MemoryMapMaxSizeMb: memoryMapMaxSizeMb,
Version: uint32(ver),
}
return volumeGrowOption, nil
}
func (ms *MasterServer) collectionInfoHandler(w http.ResponseWriter, r *http.Request) {
//get collection from request
collectionName := r.FormValue("collection")
if collectionName == "" {
writeJsonError(w, r, http.StatusBadRequest, fmt.Errorf("collection is required"))
return
}
//output details of the volumes?
detail := r.FormValue("detail") == "true"
//collect collection info
collection, ok := ms.Topo.FindCollection(collectionName)
if !ok {
writeJsonError(w, r, http.StatusBadRequest, fmt.Errorf("collection %s does not exist", collectionName))
return
}
volumeLayouts := collection.GetAllVolumeLayouts()
if detail {
//prepare the json response
all_stats := make([]map[string]interface{}, len(volumeLayouts))
for i, volumeLayout := range volumeLayouts {
volumeLayoutStats := volumeLayout.Stats()
m := make(map[string]interface{})
m["Version"] = version.Version()
m["Collection"] = collectionName
m["TotalSize"] = volumeLayoutStats.TotalSize
m["FileCount"] = volumeLayoutStats.FileCount
m["UsedSize"] = volumeLayoutStats.UsedSize
all_stats[i] = m
}
//write it
writeJsonQuiet(w, r, http.StatusOK, all_stats)
} else {
//prepare the json response
collectionStats := map[string]interface{}{
"Version": version.Version(),
"Collection": collectionName,
"TotalSize": uint64(0),
"FileCount": uint64(0),
"UsedSize": uint64(0),
"VolumeCount": uint64(0),
}
for _, volumeLayout := range volumeLayouts {
volumeLayoutStats := volumeLayout.Stats()
collectionStats["TotalSize"] = collectionStats["TotalSize"].(uint64) + volumeLayoutStats.TotalSize
collectionStats["FileCount"] = collectionStats["FileCount"].(uint64) + volumeLayoutStats.FileCount
collectionStats["UsedSize"] = collectionStats["UsedSize"].(uint64) + volumeLayoutStats.UsedSize
collectionStats["VolumeCount"] = collectionStats["VolumeCount"].(uint64) + 1
}
//write it
writeJsonQuiet(w, r, http.StatusOK, collectionStats)
}
}