Ticket #3063 (closed defect: worksforme)
Sometimes not all modules are started
| Reported by: | innocenti_jr | Owned by: | jacalvo@… |
|---|---|---|---|
| Milestone: | 2.2 | Component: | base |
| Severity: | minor | Keywords: | |
| Cc: |
Description
Intro: my server puts itself to sleep after having done a backup, wakes up in the morging and thus gets restarted quite often.
Sometimes (not clearly reproducable) the server starts just with Zentyal's core -- not starting the activated modules. No errors are logged. After another manual reboot everything gets up fine. Any clues?
I attached the relevant excerpts from zentyal.log. Note the 4 minute delay before starting the events module in 1stBoot.log@line 3.
(And maybe the ntpdate error in 2ndBoot.log@line 19 -- it fails because ntpd is already started by /etc/init.d/ntp)
(This also applies to the v2.0 series, where at times _some_ activated modules aren't started.)
Attachments
-
1stBoot.log
(1010 bytes) -
added by innocenti_jr 22 months ago.
- zentyal.log 1
-
2ndBoot.log
(5.8 KB) -
added by innocenti_jr 22 months ago.
- zentyal.log 2
-
ZentyalPackages.txt
(1.1 KB) -
added by innocenti_jr 22 months ago.
-
PsTree.txt
(629 bytes) -
added by innocenti_jr 22 months ago.
-
redis-server.log
(714 bytes) -
added by innocenti_jr 22 months ago.
Change History
Changed 22 months ago by innocenti_jr
-
attachment
1stBoot.log
added
comment:1 Changed 22 months ago by innocenti_jr
Sorry, I was about ntpd already started. On my system I already did: update-rc.d -f ntp remove to avoid such things. However, the error remains and I don't why -- but this would be another ticket. ;)
comment:2 Changed 22 months ago by jacalvo@…
- Status changed from new to assigned
No errors are logged, but, what about errors in the console? Could you try to press the "ESC" key once the logo appears to remove it and check if any error appears on screen during the failed boot?
We are going to try to reproduce it anyway, but any additional info would be very helpful.
comment:3 Changed 22 months ago by jacalvo@…
It may be also useful to know the list of your installed modules (dpkg -l | grep zentyal), maybe with that info it is easier to reproduce by us.
comment:4 Changed 22 months ago by innocenti_jr
- priority changed from normal to low
- Status changed from assigned to closed
- Resolution set to wontfix
- Severity changed from normal to minor
Well, there are no errors on the console. Other daemons run fine. I'll attach the desired package list as well as the output from "pstree -u".
Today I tried to reproduce it (ten times) myself but for the heck of it to no avail. :( So, for the meantime I close this ticket until those gremlins reappear in my server.
comment:5 Changed 22 months ago by innocenti_jr
- Status changed from closed to reopened
- Resolution wontfix deleted
Well, the gremlins keep coming back (~ 2 times/week). I attached redis' log but again, there's no error there. One can see that 4 minute gap here, too. How does Zentyal makes sure it can talk to Redis on start up? Maybe it's trying to get the list of modules from Redis to load too early?
comment:6 Changed 21 months ago by jacalvo@…
Could you try upgrading to the current packages to see if this still happens after that? Thanks!
comment:7 Changed 21 months ago by jacalvo@…
- Status changed from reopened to closed
- Resolution set to worksforme
We haven't managed to reproduce this yet, but we think probably this is fixed in the [23014] changeset, which is already released as zentyal-core 2.1.29.
Please, reopen again if you still experience this after that upgrade. Thanks!
zentyal.log 1