logrotate doesn't work exited abnormally with [1] (0 was expected)
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Fuel for OpenStack |
Opinion
|
High
|
Bartłomiej Piotrowski | ||
6.0.x |
Invalid
|
Undecided
|
Unassigned | ||
8.0.x |
Invalid
|
Undecided
|
Unassigned | ||
StackLight |
Fix Released
|
High
|
Swann Croiset | ||
0.10 |
Fix Released
|
High
|
Swann Croiset | ||
0.8 |
In Progress
|
High
|
Swann Croiset | ||
0.9 |
Fix Committed
|
High
|
Swann Croiset |
Bug Description
api: '1.0'
astute_sha: 055b2d82fe8499b
auth_required: true
build_id: 2015-04-16_21-30-10
build_number: '317'
feature_groups:
- mirantis
fuellib_sha: db5f39e96e7ab9f
fuelmain_sha: 0de2d2039e76839
nailgun_sha: 52d92c86e68602f
openstack_version: 2014.2-6.1
ostf_sha: b0991dbad159f53
production: docker
python-
release: '6.1'
Successfully deployed the following configuration:
Baremetal,
Controllers:3 Computes:47
We have found that there is no archived logs on controller nodes:
root@node-1:~# ls -lah /var/log/*-all.log*
-rw-r----- 1 syslog syslog 279M May 14 13:59 /var/log/
-rw-r----- 1 syslog syslog 356M May 14 13:59 /var/log/
-rw-r----- 1 syslog syslog 157M May 14 13:59 /var/log/
-rw-r----- 1 syslog syslog 1.8G May 14 13:59 /var/log/
-rw-r----- 1 syslog syslog 3.3G May 14 13:59 /var/log/
-rw-r----- 1 syslog syslog 718M May 14 13:59 /var/log/
For example cinder-all log hasn't been rotated:
root@node-1:~# head -1 /var/log/
<159>May 12 22:41:11 node-1 cinder-manage backend <module 'cinder.
root@node-1:~# tail -1 /var/log/
<158>May 14 14:00:28 node-1 cinder-api 192.168.0.47 - - [14/May/2015 14:00:28] "OPTIONS / HTTP/1.0" 200 422 0.003460
root@node-1:~# ls -la /var/log/cinder/
total 416764
drwxr-x--- 2 cinder adm 4096 May 12 22:41 .
drwxrwxr-x 25 root syslog 4096 May 14 12:30 ..
-rw-r--r-- 1 cinder cinder 300518822 May 14 14:00 cinder-api.log
-rw-r--r-- 1 cinder cinder 709235 May 14 14:00 cinder-backup.log
-rw-r--r-- 1 cinder cinder 23014 May 12 22:41 cinder-manage.log
-rw-r--r-- 1 cinder cinder 29438943 May 14 14:00 cinder-
-rw-r--r-- 1 cinder cinder 96045179 May 14 14:00 cinder-volume.log
from messages log file:
<13>May 14 13:50:01 node-1 logrotate: ALERT exited abnormally with [1] (0 was expected)
We can see this alert every 10 minutes
Diagnostic snapshot is here http://
no longer affects: | mos |
Changed in fuel: | |
assignee: | nobody → Bartlomiej Piotrowski (bpiotrowski) |
milestone: | none → 6.1 |
importance: | Undecided → High |
Changed in fuel: | |
status: | Incomplete → Invalid |
Changed in lma-toolchain: | |
importance: | Undecided → High |
assignee: | nobody → LMA-Toolchain Fuel Plugins (mos-lma-toolchain) |
status: | New → Confirmed |
no longer affects: | lma-toolchain/1.0 |
Changed in lma-toolchain: | |
status: | Fix Committed → Fix Released |
It's a really old iso. There were a few changes to logrotate configuration in the meantime:
https:/ /github. com/stackforge/ fuel-library/ commit/ c8702d93371e138 e8ed362cf25e959 ce960d79f3 /github. com/stackforge/ fuel-library/ commit/ 86947bb9df42b7f 3bd8625be456ffe 2dafdc50ee /github. com/stackforge/ fuel-library/ commit/ eb3e29caec4c297 9d22e46a470610e 8f6b8ac751
https:/
https:/
If you think that the issue is still present, please post here the logs from logrotate runs, because I don't think those are included in diagnostic snapshot.
1. logrotate -v /etc/logrotate.conf
2. Edit our logrotate shell wrapper to include 'set -x' before anything gets executed and attach the output of run.