mirror of
https://github.com/seaweedfs/seaweedfs.git
synced 2026-09-08 15:41:15 +02:00
Chris Lu
8398af3572
* proto: resync the java copy of filer.proto The Makefile keeps other/java/client/src/main/proto/filer.proto a verbatim copy, but AssignVolumeResponse.fsync and SubscribeMetadataResponse.flushed_ts_ns landed without it. Copy them over; no behaviour change. Claude-Session: https://claude.ai/code/session_01Fx1Hx8RqsJqHpbfbgTf4WJ * filer: route exclusive and conditional creates to the entry's ring owner CreateEntry with o_excl is a FindEntry-then-Insert. The per-path lock added for it makes that atomic only on the filer running it, and the store's insert is an upsert on every backend, so two filers both pass the existence check and both report success. mkdir(2) then succeeds twice for the same path. The same hole sits under the condition precondition, whose comment already told callers to route the key's writes to the owner filer themselves. Do it on the server instead, with the mechanism ObjectTransaction already uses: resolve the entry's ring owner and forward one hop, bounded by is_moved. The ring's membership comes from the master, so it tolerates a stale view and reassigns when a filer dies, neither of which a client's configured filer list can do. Every creator gets this — mount, S3, the filer's own HTTP surface, the Java client — not only the ones that opted in. Plain creates are upserts whoever applies them, so they stay local and pay nothing. The route key shares the S3 gateway's namespace so an object's ObjectTransaction and its CreateEntry land on the same filer's per-path lock. Claude-Session: https://claude.ai/code/session_01Fx1Hx8RqsJqHpbfbgTf4WJ
58 lines
2.2 KiB
Go
58 lines
2.2 KiB
Go
package weed_server
|
|
|
|
import (
|
|
"context"
|
|
|
|
"github.com/seaweedfs/seaweedfs/weed/glog"
|
|
"github.com/seaweedfs/seaweedfs/weed/pb"
|
|
"github.com/seaweedfs/seaweedfs/weed/s3api/s3_constants"
|
|
"github.com/seaweedfs/seaweedfs/weed/util"
|
|
)
|
|
|
|
// writeOwner returns the filer that serializes writes to key, or "" when this
|
|
// filer is the serialization point — because it owns the key, or because there
|
|
// is no ring and every filer applies locally.
|
|
//
|
|
// A ring change hands a key to its new owner before that owner has rebuilt the
|
|
// locks the prior owner still holds, so the prior owner keeps the key until the
|
|
// cooling-off window closes.
|
|
func (fs *FilerServer) writeOwner(key string) pb.ServerAddress {
|
|
if fs.filer.Dlm == nil {
|
|
return ""
|
|
}
|
|
owner := fs.filer.Dlm.LockRing.WriteOwner(key)
|
|
if owner == fs.option.Host {
|
|
return ""
|
|
}
|
|
return owner
|
|
}
|
|
|
|
// forwardToWriteOwner sends the request to key's write owner so a single filer's
|
|
// per-path lock arbitrates every writer of that key. handled=false means this
|
|
// filer is the owner and the caller should apply the request locally.
|
|
//
|
|
// An unreachable owner fails the request; it is never re-sent to another filer.
|
|
// gRPC reports a response lost in transit as Unavailable, indistinguishable from
|
|
// one the owner never saw, so a retry elsewhere could re-apply what the owner
|
|
// already committed — and an owner unreachable from here may be partitioned
|
|
// rather than down, still serving the key to everyone else. The ring hands the
|
|
// key on when the cooling-off window closes, so the outage is bounded.
|
|
func (fs *FilerServer) forwardToWriteOwner(ctx context.Context, key string, send func(owner pb.ServerAddress) error) (handled bool, err error) {
|
|
owner := fs.writeOwner(key)
|
|
if owner == "" {
|
|
return false, nil
|
|
}
|
|
if err := send(owner); err != nil {
|
|
glog.V(1).InfofCtx(ctx, "route %s to owner %s: %v", key, owner, err)
|
|
return true, err
|
|
}
|
|
return true, nil
|
|
}
|
|
|
|
// entryRouteKey is the ring key for an entry's writes. It shares the S3
|
|
// gateway's namespace so an object's ObjectTransaction and its CreateEntry
|
|
// resolve to the same owner, and land on that filer's one per-path lock.
|
|
func entryRouteKey(fullpath util.FullPath) string {
|
|
return s3_constants.ObjectWriteRouteKeyPrefix + string(fullpath)
|
|
}
|