Hi i’m using seafle server 6.0.7 on a raspberry pi v3.
I created a library on a flash key mounted through ntfs-3g. In order to format the usb stick to ext4 I transferred the data directory to a temporary place on a usb disk with the cp command.
After that process I continued adding files to the library, which is now 120Gb.
The web client and the android and windows clients work, while the seadrive does not.
It turns out that logs of seadrive shows that a block cannot be decompressed.
I then tried to perform seaf-fsck.sh revelaling a series of corrupted files.
I removed that files from the windows directory and let seafile sync.
Even after that procedure there’s a problem with a block that seaf-gc highlights but that seaf-fsck cannot repair.
As the library cannot be repaired, how can I completely reset the seafile databases in order to restart from scratch?
Thank you for your attention,
LionHe
seadrive logs. The following log repeats continuously
[01/15/17 20:33:26] sync-mgr.c(1131): Cloning repo Fotografie Seafile(5390d1bb-79ea-4840-b335-3621fffbed4a).
[01/15/17 20:33:26] sync-mgr.c(362): Repo 'Fotografie Seafile' sync state transition from 'synchronized' to 'get token'.
[01/15/17 20:33:26] [01/15/17 20:33:26] sync-mgr.c(362): Repo 'Fotografie Seafile' sync state transition from 'get token' to 'downloading'.
http-tx-mgr.c(4464): Download with HTTP sync protocol version 1.
[01/15/17 20:33:26] http-tx-mgr.c(980): Transfer repo '5390d1bb': ('normal', 'init') --> ('normal', 'check')
[01/15/17 20:33:26] http-tx-mgr.c(980): Transfer repo '5390d1bb': ('normal', 'check') --> ('normal', 'commit')
[01/15/17 20:33:26] http-tx-mgr.c(980): Transfer repo '5390d1bb': ('normal', 'commit') --> ('normal', 'fs')
[01/15/17 20:33:28] http-tx-mgr.c(3810): Bad response code for GET https:/xxxx.xxx.xxx/seafhttp/repo/5390d1bb-79ea-4840-b335-3621fffbed4a/fs-id-list/?server-head=cd82ce6953a71c0ca33d6eb7e61a76ea98bb570e&dir-only=1: 500.
[01/15/17 20:33:28] http-tx-mgr.c(4492): Failed to get fs id list for repo 5390d1bb on server https://lionhe.ddns.net.
[01/15/17 20:33:28] http-tx-mgr.c(980): Transfer repo '5390d1bb': ('normal', 'fs') --> ('error', 'finished')
[01/15/17 20:33:28] sync-mgr.c(392): Repo 'Fotografie Seafile' sync state transition from downloading to 'error': 'Error occured in download.'.
The Garbage collection script reveals that a block is damaged
./seaf-gc.sh
Starting seafserv-gc, please wait ...
[01/15/17 16:03:55] gc-core.c(440): GC version 1 repo P8Lio(f376bf26-3d40-4fe8-a5fc-6ec4d9c044cb)
[01/15/17 16:03:55] gc-core.c(313): GC started. Total block number is 440.
[01/15/17 16:03:55] gc-core.c(46): GC index size is 1024 Byte.
[01/15/17 16:03:55] gc-core.c(327): Populating index.
[01/15/17 16:03:55] gc-core.c(181): Populating index for repo f376bf26.
[01/15/17 16:03:56] gc-core.c(234): Traversed 171 commits, 440 blocks.
[01/15/17 16:03:56] gc-core.c(341): Scanning and deleting unused blocks.
[01/15/17 16:03:56] gc-core.c(364): GC finished. 440 blocks total, about 440 reachable blocks, 0 blocks are removed.
[01/15/17 16:03:56] gc-core.c(440): GC version 1 repo Fotografie Seafile(5390d1bb-79ea-4840-b335-3621fffbed4a)
[01/15/17 16:03:56] gc-core.c(313): GC started. Total block number is 126101.
[01/15/17 16:03:56] gc-core.c(46): GC index size is 63050 Byte.
[01/15/17 16:03:56] gc-core.c(327): Populating index.
[01/15/17 16:03:56] gc-core.c(181): Populating index for repo 5390d1bb.
[01/15/17 16:04:31] Empty input for zlib, invalid.
[01/15/17 16:04:31] ../../common/fs-mgr.c(1224): Failed to decompress seafile object 359c675f9175c1ed1eb97fdcaf8eb549c04ab418.
[01/15/17 16:04:31] gc-core.c(80): Failed to find file 5390d1bb-79ea-4840-b335-3621fffbed4a:359c675f9175c1ed1eb97fdcaf8eb549c04ab418.
[01/15/17 16:04:31] ../../common/commit-mgr.c(502): [comit-mgr] CommitTraverseFunc failed
[01/15/17 16:04:31] gc-core.c(234): Traversed 1 commits, 97269 blocks.
[01/15/17 16:04:31] gc-core.c(440): GC version 1 repo My Library Template(29671b35-c3af-4c5f-b3a2-97f51efa3b52)
[01/15/17 16:04:31] gc-core.c(313): GC started. Total block number is 1.
[01/15/17 16:04:31] gc-core.c(46): GC index size is 1024 Byte.
[01/15/17 16:04:31] gc-core.c(327): Populating index.
[01/15/17 16:04:31] gc-core.c(181): Populating index for repo 29671b35.
[01/15/17 16:04:31] gc-core.c(234): Traversed 2 commits, 1 blocks.
[01/15/17 16:04:31] gc-core.c(341): Scanning and deleting unused blocks.
[01/15/17 16:04:31] gc-core.c(364): GC finished. 1 blocks total, about 1 reachable blocks, 0 blocks are removed.
[01/15/17 16:04:31] gc-core.c(384): === Repos deleted by users ===
[01/15/17 16:04:31] gc-core.c(392): GC deleted repo d02c20f6.
[01/15/17 16:04:33] gc-core.c(392): GC deleted repo a01adbd1.
[01/15/17 16:04:33] gc-core.c(392): GC deleted repo 89784694.
[01/15/17 16:04:33] gc-core.c(392): GC deleted repo 625fc52f.
[01/15/17 16:04:37] gc-core.c(392): GC deleted repo 2fc56dc1.
[01/15/17 16:04:41] gc-core.c(392): GC deleted repo 10f2dee9.
[01/15/17 16:06:09] gc-core.c(456): === GC is finished ===
[01/15/17 16:06:09] gc-core.c(460): The following repos are damaged. You can run seaf-fsck to fix them.
[01/15/17 16:06:09] gc-core.c(463): 5390d1bb-79ea-4840-b335-3621fffbed4a
seafserv-gc run done
the seaf-fsck --repair output do not show anything.
./seaf-fsck.sh --repair
Starting seaf-fsck, please wait ...
[01/15/17 16:20:04] fsck.c(586): Running fsck for repo f376bf26-3d40-4fe8-a5fc-6ec4d9c044cb.
[01/15/17 16:20:04] fsck.c(413): Checking file system integrity of repo P8Lio(f376bf26)...
[01/15/17 16:20:18] fsck.c(650): Fsck finished for repo f376bf26.
[01/15/17 16:20:18] fsck.c(586): Running fsck for repo 5390d1bb-79ea-4840-b335-3621fffbed4a.
[01/15/17 16:20:18] fsck.c(413): Checking file system integrity of repo Fotografie Seafile(5390d1bb)...
[01/15/17 17:32:08] fsck.c(650): Fsck finished for repo 5390d1bb.
[01/15/17 17:32:08] fsck.c(586): Running fsck for repo 29671b35-c3af-4c5f-b3a2-97f51efa3b52.
[01/15/17 17:32:08] fsck.c(413): Checking file system integrity of repo My Library Template(29671b35)...
[01/15/17 17:32:08] fsck.c(650): Fsck finished for repo 29671b35.
seaf-fsck run done