Syncthing crashes when some folders are un-paused

Hi, my synchthing instance keeps crashing after I un-pause more than 3 folders. What log should I provide you with in order to help me figure out the problem? Syncthing v0.14.31 running on Debian 8 64bit.

Thannnk you very much!

If it crashed it would log the stacktraces explaining the crash to stdout (its normal logging location). These stacktraces would explain the crash.

How do you know that it’s crashing though? If its disconnecting, then this is expected.

I have supposed that it crashes because the service goes down.

root@SRVR:~$ service syncthing@root status
● syncthing@root.service - Syncthing - Open Source Continuous File Synchronization for root
   Loaded: loaded (/etc/systemd/system/syncthing@.service; disabled)
   Active: active (running) since Sat 2017-07-08 19:12:03 CEST; 1min 27s ago
     Docs: man:syncthing(1)
 Main PID: 18282 (syncthing)
   CGroup: /system.slice/system-syncthing.slice/syncthing@root.service
           └─18282 /usr/bin/syncthing -no-browser -no-restart -logflags=0

Jul 08 19:12:55 SRVR syncthing[18282]: [M5B2S] INFO: Access the GUI via .../
Jul 08 19:12:55 SRVR syncthing[18282]: [M5B2S] INFO: Completed initial s...)
Jul 08 19:12:55 SRVR syncthing[18282]: [M5B2S] INFO: Completed initial s...d
Jul 08 19:12:55 SRVR syncthing[18282]: [M5B2S] INFO: Completed initial s...p
Jul 08 19:12:55 SRVR syncthing[18282]: [M5B2S] INFO: Completed initial s...x
Jul 08 19:12:57 SRVR syncthing[18282]: [M5B2S] INFO: Completed initial s...a
Jul 08 19:12:58 SRVR syncthing[18282]: [M5B2S] INFO: Completed initial s...s
Jul 08 19:12:58 SRVR syncthing[18282]: [M5B2S] INFO: Completed initial s...s
Jul 08 19:13:13 SRVR syncthing[18282]: [M5B2S] INFO: Established secure ...)
Jul 08 19:13:13 SRVR syncthing[18282]: [M5B2S] INFO: Device [DEVICE ID xxx]..."
Hint: Some lines were ellipsized, use -l to show in full.
root@SRVR:~$ service syncthing@root status
● syncthing@root.service - Syncthing - Open Source Continuous File Synchronization for root
   Loaded: loaded (/etc/systemd/system/syncthing@.service; disabled)
   Active: inactive (dead) since Sat 2017-07-08 19:15:46 CEST; 6s ago
     Docs: man:syncthing(1)
  Process: 18282 ExecStart=/usr/bin/syncthing -no-browser -no-restart -logflags=0 (code=exited, status=2)
 Main PID: 18282 (code=exited, status=2)

Jul 08 19:15:46 SRVR syncthing[18282]: /home/buildslave/workspace/syncth...f
Jul 08 19:15:46 SRVR syncthing[18282]: created by github.com/syncthing/s...n
Jul 08 19:15:46 SRVR syncthing[18282]: /home/buildslave/workspace/syncth...7
Jul 08 19:15:46 SRVR syncthing[18282]: goroutine 1749 [chan receive]:
Jul 08 19:15:46 SRVR syncthing[18282]: github.com/syncthing/syncthing/li...)
Jul 08 19:15:46 SRVR syncthing[18282]: /home/buildslave/workspace/syncth...f
Jul 08 19:15:46 SRVR syncthing[18282]: github.com/syncthing/syncthing/li...)
Jul 08 19:15:46 SRVR syncthing[18282]: /home/buildslave/workspace/syncth...5
Jul 08 19:15:46 SRVR syncthing[18282]: created by github.com/syncthing/s...n
Jul 08 19:15:46 SRVR syncthing[18282]: /home/buildslave/workspace/syncth...1
Hint: Some lines were ellipsized, use -l to show in full.

Sorry for double post. Maybe this is more helpful:

root@GC01SRVR:/var/log$ tail -n1000 /var/log/syslog | grep syncthing
Jul  8 19:34:10 SRVR syncthing[18541]: panic: read /root/.config/syncthing/index-v0.14.0.db/085861.ldb: input/output error

Yep, the problem in with this .db file:

root@SRVR:/var/log$ cat /var/log/syslog | grep "panic"
Jul  8 19:15:46 SRVR syncthing[18282]: panic: read /root/.config/syncthing/index-v0.14.0.db/085861.ldb: input/output error
Jul  8 19:22:29 SRVR syncthing[18362]: panic: read /root/.config/syncthing/index-v0.14.0.db/085861.ldb: input/output error
Jul  8 19:34:10 SRVR syncthing[18541]: panic: read /root/.config/syncthing/index-v0.14.0.db/085861.ldb: input/output error
Jul  8 19:54:22 SRVR syncthing[18755]: panic: read /root/.config/syncthing/index-v0.14.0.db/085861.ldb: input/output error

Any suggestion on what to do?

Thank you!

The disk on which syncthing database lives is not healthy

Yes, I am planning to change it. Any idea on how to make Syncthing start again?

It should start when it does not get an I/O error from the database. Hopefully you can just move it to a healthy disk, and it should recover what it can from the existing database. If that still fails for some reason, syncthing -reset-database or just delete the index-something.db directory. Ideally, be in sync when you do this, as whatever you have locally will be considered the latest and greatest when you delete the database.

Thank you very much to both of you @calmh and @AudriusButkevicius! Taking into consideration what @calmh said:

  • since my configuration is basically server-client (client is my laptop and server is my server, any change on client is reproduced on server)
  • since the problem occurred on the server
  • since the two machines were not synced

if I reset the database both on client and server, will the changes on the client prevail?

If you reset the database on both sides and the files differ it will look like a simultaneous modification and generate a conflict copy.

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.