SeaweedFS - Package Updates

[Package Updates]

[1.10.0]

• Update seaweedfs to 4.33
• Full Changelog
• fix(s3api): accept HTTP-date conditionals by @7y-9 in #9863
• fix(s3api): drop ancestor directory markers from prefixed ListObjectVersions by @chrislusf in #9885
• fix(remote): correct content and permissions when syncing/caching remote objects by @chrislusf in #9879
• fix(filer.sync): replicate a rename as an atomic move, not a no-op update by @chrislusf in #9895
• feat(admin): export full cluster volume list as JSON by @chrislusf in #9876
• feat(filer): object size distribution metric and dashboard panels by @chrislusf in #9902
• fix(seaweed-volume): stop EC shard deletion from phantom .dat on restart by @chrislusf in #9874
• add elastic8 filer store for Elasticsearch 8 by @chrislusf in #9916
• fix(ec): don't mix EC shards from different encode runs by @chrislusf in #9880
• Fix filer metadata-replay OOM under mount reconnect storms by @chrislusf in #9901

[Package Updates]

[1.11.0]

• Update seaweedfs to 4.34
• Full Changelog
• s3: export per-bucket quota and read-only state metrics by @chrislusf in #9923
• s3: validate copy source path segments by @chrislusf in #9929
• fix(util): ignore comment only sql input by @7y-9 in #9933
• shell: volume.tier.move can move volumes between data centers by @chrislusf in #9925
• admin: configure maintenance tasks via admin.toml by @chrislusf in #9926
• feat(storage): sweep leftover empty EC .dat stubs on volume server startup by @chrislusf in #9927
• fix(storage): never let an empty .dat delete healthy distributed EC shards by @chrislusf in #9930
• fix(replication): verify-before-destroy in VolumeCopy, check.disk, and over-replication trim by @chrislusf in #9943
• fix(vacuum): crash-safe compaction commit with a durable .cpc marker, fsync-before-rename, and a reload fence by @chrislusf in #9944
• metrics: overhaul Grafana dashboard for full metric coverage by @chrislusf in #9956

[Package Updates]

[1.12.0]

• Update seaweedfs to 4.35
• Full Changelog
• fix(command): preserve fuse option after writers by @7y-9 in #9972
• mount: tolerance-window write pattern detection for concurrent writeback by @chrislusf in #9984
• feat(mount): attach Content-MD5 to chunk uploads by @kisow in #10016
• mq(kafka): don't drop an existing topic when auto-create races by @chrislusf in #9998
• feat: add Prometheus metrics for replication operations by @rushikesh90 in #10006
• feat: add Prometheus metric for volume creation operations by @rushikesh90 in #10026
• volume: remove ec.bitrotChecksum and ec.bitrotBlockSizeMB flags by @chrislusf in #10000
• s3: bound streaming remote-cache wait so large cold GetObject returns 503, not a hang by @chrislusf in #9988
• fix(s3): verify SigV2 using percent-encoded path for Unicode object keys by @sergey-zinchenko in #10022
• helm: reject emptyDir for volume idx, and rebuild a missing idx on restart by @chrislusf in #10005

[Package Updates]

[1.13.0]

• Update seaweedfs to 4.36
• Full Changelog
• s3api: sort repeated SigV4 query values @7y-9 (#10031)
• s3api: reject malformed Range offsets @guankai (#10034)
• s3: propagate IAM changes from standalone weed s3 to peer pods @chrislusf (#10095)
• iceberg: support multi-table transaction commit @chrislusf (#10066)
• iceberg: support views @chrislusf (#10069)
• sts: authorize AssumeRole by the role's trust policy @chrislusf (#10097)
• feat: support marking volumes by collection @7y-9 (#9585)
• mount: don't hang close() when a writer is killed during flush @chrislusf (#10090)
• postgres: resolve startup message length type mismatch and uint underflow OOM risk @198wmj (#10065)
• util/http: lazily init the global HTTP client to fix admin metrics nil panic @chrislusf (#10044)

[Package Updates]

[1.14.0]

• Update seaweedfs to 4.37
• Full Changelog
• fix (Volume [Rust]): stream copy_file and volume_incremental_copy instead of buffering the whole file in memory by @justadri in #10110
• fix(volume): serve reads from remote after tier upload (Rust) by @chrislusf in #10112
• fix(volume): make tier-down crash-safe and serve from local (Rust) by @chrislusf in #10113
• volume server: route VolumeMarkReadonly to raft leader by @jk2lx in #10120
• fix: apply collectionPattern during detection in volume.fix.replication by @qzhello in #10115
• ec: recover EC shards whose .ecx index lives only on a peer server by @chrislusf in #10108
• feat(metrics): Add EC rebuild/reconstruct Prometheus metrics by @rushikesh90 in #10124
• fix(master): shed assign load when volume growth is already in flight by @chrislusf in #10121
• s3: support AWS object form for bucket policy Principal, add NotPrincipal by @chrislusf in #10125
• fix(filer.sync.verify): sort listings client-side before merge by @chrislusf in #10117

[Package Updates]

[1.15.0]

• Update seaweedfs to 4.38
• Full Changelog
• #10130 Fix scrubbing of deleted needles on EC volumes
• #10186 return IncompleteBody instead of 500 for truncated PUT bodies
• #10192 embedded IAM inline policy honors prefix-scoped resources
• #10203 stream offloaded metadata-log entries to fix concurrent-write OOM
• #10185 drop stale volume-location cache on under-replication
• #10229 bound intra-cluster HTTP so an unresponsive peer can't hang reads/writes
• #10214 paginate ListBuckets and serve from a bucket owner index
• #10224 enforce bucket quota on logical size + read-only state in Admin UI
• #10204 add ec.shard.unmount command
• #10236 FULL_OBJECT checksumming for CRC

[Package Updates]

[1.16.0]

• Update seaweedfs to 4.39
• Full Changelog
• async, buffered writes in VolumeEcShardsCopy (#10237)
• fix orphan purge against the rust volume server (#10289)
• keep tier-uploaded volume reporting to master after volume.tier.upload (#10259)
• require JWT authorization on TUS upload endpoints (#10249)
• fail over routed object writes when the owner filer is unreachable (#10251)
• keep listing when empty directories fill the listing window (#10280)
• verify SigV4 against each plausible reverse-proxy host (#10284)
• surface OIDC groups and roles into the STS session request context for resource-policy ABAC (#10263)
• re-assign to a live volume when a write can't land (#10239)
• merge filer service annotations to avoid duplicate keys in Helm (#10293)

[Package Updates]

[1.17.0]

• Update seaweedfs to 4.40
• Full Changelog
• s3: enforce role trust policy on direct OIDC bearer authentication by @chrislusf in #10302
• s3: accept raw semicolons in query strings by @chrislusf in #10305
• s3: return NoSuchKey/NoSuchBucket for a missing CopyObject source by @chrislusf in #10332
• Fix PutObjectAcl writing back to the wrong object for nested keys by @chrislusf in #10333
• fix(filer.backup): prevent silent backup data loss on transient not-found by @kisow in #10295
• filer: harden TUS session authorization against cross-prefix access by @chrislusf in #10315
• ec: remove .ecsum sidecars on destroy / shard delete; align Go and Rust cleanup by @franchb in #10307
• Introduce weed shell command ec.check.replication. by @chrislusf in #10328
• fix(volume): reject overflowing needle ID deltas by @7y-9 in #10342
• mount: surface ENOSPC instead of endless waiting in #10341

[Package Updates]

[1.18.0]

• Update seaweedfs to 4.41
• Full Changelog
• S3: fold large chunk lists into manifest chunks on the direct write path by @chrislusf in #10383
• s3: stop listing prefixes whose objects are all delete-marked by @chrislusf in #10419
• s3: require a bucket-policy action to write a bucket policy by @chrislusf in #10444
• fix(filer): stop skipping recent unflushed events on metadata subscription gaps by @kisow in #10501
• fix(redis2/redis3): support separate sentinel auth credentials by @TJDawson10 in #10412
• Fix volume.merge corrupting every needle it copies by @chrislusf in #10565
• fix(volume): persist the leveldb needle map watermark at batch boundaries by @SuperMohit in #10557
• fix(master): bump seaweedfs/raft to v1.2.0 for the snapshot race by @chrislusf in #10498
• mount: support mounting on Windows through WinFsp by @chrislusf in #10536
• azure: authenticate with Entra ID instead of a storage account key by @chrislusf in #10456

[Package Updates]

[1.19.0]

• Update seaweedfs to 4.42
• Full Changelog
• storage: fix corrupted leveldb detection in DoOffsetLoading (#10650)
• mount: fix four readdir pagination bugs (#10624)
• s3api: fix ListObjectsV2 dropping objects under a partial prefix (#10656)
• s3: a key deleted after enabling versioning must leave the listing (#10684)
• filer: do not sweep children when deleting a folder non-recursively (#10782)
• filer: restore a folder that received an entry while it was deleted (#10783)
• s3: add the RenameObject endpoint (#10659)
• s3: option to disable bucket auto-creation on upload (#10759)
• throughput limits for replicate, EC shard, and worker-driven moves (#10749)
• grpc: optional client_cert/client_key for outgoing mTLS connections (#10747)

[Package Updates]

[1.20.0]

• Update seaweedfs to 4.43
• Full Changelog
• s3: stop unrouted bucket subresources from being answered with a listing by @chrislusf in #10814
• s3: place multipart part chunks by the destination object's storage rule by @chrislusf in #10845
• filer: use bind variables for request-controlled values in the arangodb store by @chrislusf in #10795
• filer: stop remote.unmount from deleting the remote objects by @Koalk in #10811
• fix(filer): bound aggregated metadata reads by peer watermarks by @kisow in #10803
• Recover from a dead volume server on the mount read path by @chrislusf in #10798
• Re-look-up a chunk's locations as soon as they all fail by @chrislusf in #10800
• volume: forward fsync=true to replicas in ReplicatedWrite by @sunchao6106 in #10805
• Lance catalog, and a Rust plugin worker to maintain it by @chrislusf in #10841
• Fix what the Go 1.26 language bump breaks by @chrislusf in #10794