Samba groups disappear

[nardial@…]

Hi,

my Samba Groups disappear every day and the redis server is crashing somehow. When I stop zental-users the redis process is still running:

130       6517  0.0  0.0   9832   716 ?        Ss   Jun18   0:07 /usr/bin/redis-server /var/lib/zentyal-usercorner/conf/redis.conf
ebox     10404  0.0  0.2  13660  5524 ?        Ss   Jun18   0:09 /usr/bin/redis-server /var/lib/zentyal/conf/redis.conf
redis    11189  0.0  0.0   9832  1280 ?        Ss   Jun18   0:07 /usr/bin/redis-server /etc/redis/redis.conf

So, I have to kill the process (here 10404) and start zentyal users again. Then the groups reappear in the webinterface. If I do not do this regularly, the groups disappear in all samba shares and all users are getting crazy ;)

zentyal.log:

2012/06/19 07:23:45 INFO> Redis.pm:766 EBox::Config::Redis::_initRedis - Starting redis server
2012/06/19 07:23:47 INFO> Redis.pm:766 EBox::Config::Redis::_initRedis - Starting redis server
2012/06/19 07:23:49 INFO> Redis.pm:766 EBox::Config::Redis::_initRedis - Starting redis server
2012/06/19 07:24:24 INFO> Redis.pm:766 EBox::Config::Redis::_initRedis - Starting redis server
2012/06/19 07:24:26 INFO> Redis.pm:766 EBox::Config::Redis::_initRedis - Starting redis server
2012/06/19 07:24:29 INFO> Redis.pm:766 EBox::Config::Redis::_initRedis - Starting redis server
2012/06/19 07:29:22 INFO> Redis.pm:766 EBox::Config::Redis::_initRedis - Starting redis server
2012/06/19 07:31:01 INFO> Redis.pm:766 EBox::Config::Redis::_initRedis - Starting redis server

/var/log/zentyal/redis-server.log:

[16977] 19 Jun 07:31:01 # Opening TCP port: bind: Address already in use

[16978] 19 Jun 07:31:01 # Opening TCP port: bind: Address already in use

[16979] 19 Jun 07:31:01 # Opening TCP port: bind: Address already in use

[16980] 19 Jun 07:31:01 # Opening TCP port: bind: Address already in use

[16981] 19 Jun 07:31:01 # Opening TCP port: bind: Address already in use

[16982] 19 Jun 07:31:02 # Opening TCP port: bind: Address already in use

[16983] 19 Jun 07:31:02 # Opening TCP port: bind: Address already in use

[16984] 19 Jun 07:31:02 # Opening TCP port: bind: Address already in use

/var/log/syslog:

Jun 19 07:31:01 XXX init: ebox.redis main process (16979) terminated with status 1
Jun 19 07:31:01 XXX init: ebox.redis main process ended, respawning
Jun 19 07:31:01 XXX slapd[4321]: connection_read(48): no connection!
Jun 19 07:31:01 XXX slapd[4321]: connection_read(48): no connection!

zentyal-users: 2.2.6
zentyal-core: 2.2.7
redis-server: 2:1.2.0-1

[jamor@…]

Hello Nardial,

first the redis has nothing to do with this whis soptting/restarting the user modules so we can forget about it.

The groups dissaparition is caused likely by slapd stopping or losing the connection. I am assuming you ae using standalone server. Can you look to zentyal.log and syslog to find more clues about slapd?.

If you don't find nothing in the logs or more clues, I cannot offer much because this seems that something particular went wrong in your system. Maybe doin a backup, reinstalling users and groups ('sudo /usr/share/zentyal-users/reinstall') with the same DN and then restore the backup?.

Regards,

Javier

[nardial@…]

Hi Javier,

i can only find these log messages in syslog:

Jun 19 00:01:01 XXX init: ebox.redis main process (598) terminated with status 1
Jun 19 00:01:01 XXX init: ebox.redis main process ended, respawning
Jun 19 00:01:01 XXX slapd[14957]: connection_read(13): no connection!
Jun 19 00:01:01 XXX slapd[14957]: connection_read(13): no connection!

and

Jun 19 00:30:28 XXX slapd[14957]: <= bdb_equality_candidates: (mail) not indexed
Jun 19 00:30:28 XXX slapd[14957]: <= bdb_equality_candidates: (dc) not indexed
Jun 19 00:30:28 XXX slapd[14957]: <= bdb_equality_candidates: (mail) not indexed
Jun 19 00:30:28 XXX slapd[14957]: <= bdb_equality_candidates: (mail) not indexed

I thought about reinstalling the users module myself, but did not dare to do it. I'll take a snapshot of the VM and hopefully the internal zentyal backup can be restored. Do I simply have to do a full backup and restore this after reinstalling the zentyal-users module or just backup the configuration/settings?

The server has been upgraded from a late beta version

nardial

[jamor@…]

Hello,

the LDAP data is in the configuration backup. So not necessary to do a full backup.

For you information when you do a backup with the backup module a configuration backup is done automatically and included.

[jamor@…]

I close this until more feedback is given, please reopen if needed.

[nardial@…]

Hi!

Today it stopped working again. I reinstalled it on Friday and it worked on Sat/Sun?.

What is more important today is, that all my zarafa data / stores are totally lost. Where is this data gone?

[nardial@…]

Due to the fact that I had problems with my zarafa backup, I have lost quite a lot of data. Why did nobody mention, that there might be a problem with zarafa accounts, when I reinstall the user module? Restarting the users module is not so much a problem like losing mail accounts....

[jamor@…]

Sorry I don't know you have zarafa enabled. The zarafa mails are stored in the mysql database. I think that after the restore new storages for the accounts has been created, maybe you can copy the data from the old storage to the new ones?.

Anyway this section is intended for bug reporting, if we have an operation or or configuration problem the forums (  http://forum.zentyal.org/ ) is a more appropriate place. Why not do you start a thread there?.

Regards,

Javier