LizardFS master 3.10.2 crash

[andypl78]

[4581135.415771] mfsmaster[13210]: segfault at 0 ip 00000000004fa3b7 sp 00007ffc9fb8f1f0 error 4 in mfsmaster[400000+15d000]

Setup:

• Debian Wheezy;
• lizardfs-master 3.10.2 from repo deb http://packages.lizardfs.com/debian/wheezy wheezy main

[blink69]

Please send us more details regarding that,

• numbers of files,
• numbers of chunks,
• numbers of connected chunkservers, clients
  also syslog from before this crash dump will also help.

[blink69]

patch for that is in our CR http://cr.skytechnology.pl:8081/#/c/2720/

[andypl78]

When can we expect a new version in the repo?

My setup (all on version 3.10.2) :

• master x1
• shadow x1
• metalogger x1
• chunks x2
• clients x22

syslog:

Sep 26 09:17:01 lfs01 /USR/SBIN/CRON[26509]: (root) CMD (   cd / && run-parts --report /etc/cron.hourly)
Sep 26 09:47:23 lfs01 mfsmaster[13210]: structure check loop
Sep 26 10:17:01 lfs01 /USR/SBIN/CRON[14730]: (root) CMD (   cd / && run-parts --report /etc/cron.hourly)
Sep 26 11:17:01 lfs01 /USR/SBIN/CRON[2936]: (root) CMD (   cd / && run-parts --report /etc/cron.hourly)
Sep 26 12:17:01 lfs01 /USR/SBIN/CRON[23580]: (root) CMD (   cd / && run-parts --report /etc/cron.hourly)
Sep 26 13:17:01 lfs01 /USR/SBIN/CRON[11799]: (root) CMD (   cd / && run-parts --report /etc/cron.hourly)
Sep 26 13:47:37 lfs01 mfsmaster[13210]: structure check loop
Sep 26 14:17:01 lfs01 /USR/SBIN/CRON[32448]: (root) CMD (   cd / && run-parts --report /etc/cron.hourly)
Sep 26 15:17:01 lfs01 /USR/SBIN/CRON[20667]: (root) CMD (   cd / && run-parts --report /etc/cron.hourly)
Sep 26 16:17:01 lfs01 /USR/SBIN/CRON[8887]: (root) CMD (   cd / && run-parts --report /etc/cron.hourly)
Sep 26 16:42:54 lfs01 kernel: [4581135.415771] mfsmaster[13210]: segfault at 0 ip 00000000004fa3b7 sp 00007ffc9fb8f1f0 error 4 in mfsmaster[400000+15d000]
Sep 26 16:42:54 lfs01 mfsmount[25736]: master: connection lost

[blink69]

In next few days we will publish 3.10.4 with all fixes.

[viktor-zhuromskyy]

Same thing here.
lizardfs-master crashes once a day on average load.

[DrProfi]

Confirm problem with 3.10.4

[3780056.793786] mfsmaster[19333]: segfault at 7fffc4c68ff8 ip 00000000004db2b5 sp 00007fffc4c69000 error 6 in mfsmaster[400000+13a000]

[3789174.107720] mfsmaster[7194]: segfault at 7ffe06e40ff8 ip 00000000004db2b2 sp 00007ffe06e41000 error 6 in mfsmaster[400000+13a000]

From repo for ubuntu 14.04

[blink69]

This issue is related to bug in libjudy, see issue #496

[psarna]

#496 is already closed with e77895a