chrislu
2a4a9c6343
log possible racing condition
fix https://github.com/seaweedfs/seaweedfs/issues/5412
9 months ago
Konstantin Lebedev
5189a09de0
[volume] Reduce the number of buffers for uploading one chunk ( #5458 )
9 months ago
skycope
6dae685f9c
fix append disorder ( #5492 )
9 months ago
steve.wei
67ead9b18f
fix(volume.fix.replication): adjust volume count, not free volume count ( #5479 )
9 months ago
Henco Appel
5c8e6014ba
fix: filer authenticate with with volume server ( #5480 )
9 months ago
clonefetch
9e07a87fcb
chore: fix function names in comment ( #5478 )
9 months ago
Konstantin Lebedev
3e25ed1b11
[s3] add s3 pass test_multipart_upload_size_too_small ( #5475 )
* add s3 pass test_multipart_upload_size_too_small
* refactor metric names
* return ErrNoSuchUpload if empty parts
* fix test
9 months ago
Konstantin Lebedev
35cba720a5
[s3] add s3 pass test_multipart_upload ( #5474 )
add s3 pass test_multipart_upload
9 months ago
Konstantin Lebedev
d42a04cceb
[s3] fix s3 test_multipart_resend_first_finishes_last ( #5471 )
* try fix s3 test
https://github.com/seaweedfs/seaweedfs/pull/5466
* add error handler metrics
* refactor
* refactor multipartExt
* delete bad entry parts
9 months ago
Nikita Borzykh
7aa25c1137
fix: correctly form partNumber from new entry.Name format in "listObjectParts" ( #5470 )
9 months ago
skycope
b19c9847c6
fix completed multiupload lost data ( #5460 )
If there are putObjectPart requests with the same uploadId during
completeMultiPart, it can result in data loss. putObjectPart requests
might be due to timeout retries.
Co-authored-by: Yang Wang <yangwang@weride.ai>
9 months ago
chrislu
ccdd9cd8de
refactor
9 months ago
chrislu
f37c0d0d7a
comment out println
9 months ago
Konstantin Lebedev
d5d8b8e2ae
fix panic at isAllWritable ( #5457 )
fix panic
https://github.com/seaweedfs/seaweedfs/issues/5456
9 months ago
Henco Appel
2a88da4de7
fix: always close volume file ( #4530 ) ( #5459 )
If sync fails then close is never called. We should always be calling
close on the file.
Co-authored-by: Hendrik Appel <happel@europe.altair.com>
9 months ago
chrislu
d935f70e3c
shutdown follower when leader stops
9 months ago
chrislu
f07875e8e1
send flush message to follower before shutting down logBuffer
9 months ago
Nikita Borzykh
df9031e19a
fix: ensure complete gRPC stream consumption in doListFilerEntries to prevent context canceled errors. ( #5441 )
9 months ago
chrislu
e568e742c9
renaming
9 months ago
chrislu
47975e75bf
purge old code
9 months ago
chrislu
237e68059d
shutdown follower if no clients, not only publishers
to avoid overhead when a publisher repeatedly send one message and close
9 months ago
chrislu
66a878af39
add control message to data message
9 months ago
chrislu
546ae87c39
send is_close message to broker
9 months ago
chrislu
c9df613b6b
add publisher name for debugging
9 months ago
chrislu
ca4f89a6f6
simplify
9 months ago
chrislu
0d55bddc75
logs
9 months ago
Nikita Borzykh
829c0da381
fix: adjust condition in prefixFilterEntries prevent infinite loop ( #5440 )
9 months ago
Konstantin Lebedev
9c1e0f5811
[master] grow volumes if no writable volumes in current dataCenter ( #5434 )
* grow volumes if no writable volumes in current dataCenter
https://github.com/seaweedfs/seaweedfs/issues/3886
* fix tests with volume grow
* automatic volume grow one volume
* add ErrorChunkAssign metrics
9 months ago
Konstantin Lebedev
df40908e57
fix panic 5435 ( #5436 )
9 months ago
chrislu
e3c8be7f89
LocalPartition shutdown
9 months ago
chrislu
ab9fc0ad70
simplify
9 months ago
chrislu
7463058299
change visibility
9 months ago
chrislu
dab5454332
publish and send to follower
9 months ago
chrislu
f249f831b1
MaybeShutdownLocalPartition refactor
9 months ago
chrislu
2a0ab7e90d
fix
9 months ago
chrislu
059a120708
refactor
9 months ago
chrislu
7d87c1d2bb
refactor
9 months ago
chrislu
6aa804b368
lock instead of RLock, to prevent racing condition
https://github.com/seaweedfs/seaweedfs/discussions/5432#discussioncomment-8933608
9 months ago
chrislu
5cc94a05b9
separate goroutine to send ack to publisher
9 months ago
chrislu
17806cde2a
separate context
If using the same context, the downstream grpc would be cancelled when the parent grpc is cancelled.
9 months ago
chrislu
f67c8309d8
rename
9 months ago
chrislu
7d69552974
log server list
9 months ago
chrislu
36fec34c47
print only adapted url
fix https://github.com/seaweedfs/seaweedfs/issues/5424
10 months ago
chrislu
aed6fdffd2
fix nil volume data backend
fix https://github.com/seaweedfs/seaweedfs/issues/5417
10 months ago
chrislu
d6a066495b
adjust logs
10 months ago
chrislu
1b1b96ea1e
close local partition if no more publishers and subscribers
10 months ago
Nikita Borzykh
9ca871af53
Move ListDirectoryEntries logic to ListDirectoryPrefixedEntries in etcd meta storage backend ( #5416 )
Move ListDirectoryEntries logic to ListDirectoryPrefixedEntries
10 months ago
chrislu
4f5c4c3388
refactor
10 months ago
chrislu
d558a58946
refactor
10 months ago
chrislu
205d2285f3
setup follower by publisher
* the subscriber would getOrGen a local partition and wait
* the publisher would getOrGen a local partition. If localPartition follower is not setup, and init message has follower info, it would create followers based on init.Messages.
10 months ago