Hi all,
I have a problem with logrotate and I don't know why. In /var/log/messages appears these errors:
Apr 8 15:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 16:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 17:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 18:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 19:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 20:01:02 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 21:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 22:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 23:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1]
It seems the problem if with this logrotate.conf (executed every hour):
create dateext
/srv/www2/logs/viewer.log { size 512k create 0640 www www missingok rotate 7 compress delaycompress copytruncate notifempty }
Doing a debug:
[root@www02 ~]# logrotate /data/config/etc/logrotate.conf -dv reading config file /data/config/etc/logrotate.conf reading config info for /nsm/moloch/logs/viewer.log
Handling 1 logs
rotating pattern: /srv/www2/logs/viewer.log 524288 bytes (7 rotations) empty log files are not rotated, old logs are removed considering log /srv/www2/logs/viewer.log log needs rotating rotating log /srv/www2/logs/viewer.log, log->rotateCount is 7 dateext suffix '-20140409' glob pattern '-[0-9][0-9][0-9][0-9][0-9][0-9][0-9][0-9]' glob finding logs to compress failed destination /srv/www2/logs/viewer.log-20140409 already exists, skipping rotation
According to this kb from redhat's website:
https://access.redhat.com/site/solutions/39006
problem could be selinux. But SElinux is disabled:
[root@www02 ~]# sestatus SELinux status: disabled [root@www02 ~]#
Any idea?? Could be a bug??
Host is a CentOS 6.5 x86_64 fully patched.
Thanks.
From: C. L. Martinez carlopmart@gmail.com
destination /srv/www2/logs/viewer.log-20140409 already exists, skipping rotation
I'd think logrotate does not expect to find an already existing "half-rotated" file? Maybe this file is supposed to be compressed '.gz'? Can you show "ls -l /srv/www2/logs/viewer.log*" ?
JD
Logrotate, by default, runs daily (not hourly) and renames the log files that it rotates by appending to them the current date. In your case, you are running it hourly, so it only runs successfully only once a day. The second time that it will attempt to run in the same day, it will find that there already exists a rotated file with the appended date for that same day, and fail.
So, either run it daily, or change the renaming scheme.
On Wed, Apr 9, 2014 at 9:45 AM, C. L. Martinez carlopmart@gmail.com wrote:
Hi all,
I have a problem with logrotate and I don't know why. In /var/log/messages appears these errors:
Apr 8 15:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 16:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 17:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 18:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 19:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 20:01:02 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 21:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 22:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 23:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1]
It seems the problem if with this logrotate.conf (executed every hour):
create dateext
/srv/www2/logs/viewer.log { size 512k create 0640 www www missingok rotate 7 compress delaycompress copytruncate notifempty }
Doing a debug:
[root@www02 ~]# logrotate /data/config/etc/logrotate.conf -dv reading config file /data/config/etc/logrotate.conf reading config info for /nsm/moloch/logs/viewer.log
Handling 1 logs
rotating pattern: /srv/www2/logs/viewer.log 524288 bytes (7 rotations) empty log files are not rotated, old logs are removed considering log /srv/www2/logs/viewer.log log needs rotating rotating log /srv/www2/logs/viewer.log, log->rotateCount is 7 dateext suffix '-20140409' glob pattern '-[0-9][0-9][0-9][0-9][0-9][0-9][0-9][0-9]' glob finding logs to compress failed destination /srv/www2/logs/viewer.log-20140409 already exists, skipping rotation
According to this kb from redhat's website:
https://access.redhat.com/site/solutions/39006
problem could be selinux. But SElinux is disabled:
[root@www02 ~]# sestatus SELinux status: disabled [root@www02 ~]#
Any idea?? Could be a bug??
Host is a CentOS 6.5 x86_64 fully patched.
Thanks. _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
http://svn.fedorahosted.org/svn/logrotate/tags/r3-8-5/logrotate.8 shows "hourly" as a .conf option was added to 3.8.5...
current centos 6 version is logrotate-3.7.8-17.el6.x86_64
centos6$ man logrotate: dateformat format_string Specify the extension for dateext using the notation similar to strftime(3) function. Only %Y %m %d and %s specifiers are allowed.
in your .conf try: dateformat -%s
On Wed, Apr 9, 2014 at 2:45 AM, C. L. Martinez carlopmart@gmail.com wrote:
Hi all,
I have a problem with logrotate and I don't know why. In /var/log/messages appears these errors:
Apr 8 15:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 16:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 17:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 18:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 19:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 20:01:02 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 21:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 22:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 23:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1]
It seems the problem if with this logrotate.conf (executed every hour):
create dateext
/srv/www2/logs/viewer.log { size 512k create 0640 www www missingok rotate 7 compress delaycompress copytruncate notifempty }
Doing a debug:
[root@www02 ~]# logrotate /data/config/etc/logrotate.conf -dv reading config file /data/config/etc/logrotate.conf reading config info for /nsm/moloch/logs/viewer.log
Handling 1 logs
rotating pattern: /srv/www2/logs/viewer.log 524288 bytes (7 rotations) empty log files are not rotated, old logs are removed considering log /srv/www2/logs/viewer.log log needs rotating rotating log /srv/www2/logs/viewer.log, log->rotateCount is 7 dateext suffix '-20140409' glob pattern '-[0-9][0-9][0-9][0-9][0-9][0-9][0-9][0-9]' glob finding logs to compress failed destination /srv/www2/logs/viewer.log-20140409 already exists, skipping rotation
According to this kb from redhat's website:
https://access.redhat.com/site/solutions/39006
problem could be selinux. But SElinux is disabled:
[root@www02 ~]# sestatus SELinux status: disabled [root@www02 ~]#
Any idea?? Could be a bug??
Host is a CentOS 6.5 x86_64 fully patched.
Thanks. _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
Yep, Marios and Steven are right ... I have changed dateformat for one log file and I have done another test changing time rotations ... And all works ok.
Many thanks for your help.
On Fri, Apr 11, 2014 at 12:09 AM, Steven Tardy sjt5atra@gmail.com wrote:
http://svn.fedorahosted.org/svn/logrotate/tags/r3-8-5/logrotate.8 shows "hourly" as a .conf option was added to 3.8.5...
current centos 6 version is logrotate-3.7.8-17.el6.x86_64
centos6$ man logrotate: dateformat format_string Specify the extension for dateext using the notation similar to strftime(3) function. Only %Y %m %d and %s specifiers are allowed.
in your .conf try: dateformat -%s
On Wed, Apr 9, 2014 at 2:45 AM, C. L. Martinez carlopmart@gmail.com wrote:
Hi all,
I have a problem with logrotate and I don't know why. In /var/log/messages appears these errors:
Apr 8 15:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 16:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 17:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 18:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 19:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 20:01:02 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 21:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 22:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1] Apr 8 23:01:01 plzfnsm02 logrotate: ALERT exited abnormally with [1]
It seems the problem if with this logrotate.conf (executed every hour):
create dateext
/srv/www2/logs/viewer.log { size 512k create 0640 www www missingok rotate 7 compress delaycompress copytruncate notifempty }
Doing a debug:
[root@www02 ~]# logrotate /data/config/etc/logrotate.conf -dv reading config file /data/config/etc/logrotate.conf reading config info for /nsm/moloch/logs/viewer.log
Handling 1 logs
rotating pattern: /srv/www2/logs/viewer.log 524288 bytes (7 rotations) empty log files are not rotated, old logs are removed considering log /srv/www2/logs/viewer.log log needs rotating rotating log /srv/www2/logs/viewer.log, log->rotateCount is 7 dateext suffix '-20140409' glob pattern '-[0-9][0-9][0-9][0-9][0-9][0-9][0-9][0-9]' glob finding logs to compress failed destination /srv/www2/logs/viewer.log-20140409 already exists, skipping rotation
According to this kb from redhat's website:
https://access.redhat.com/site/solutions/39006
problem could be selinux. But SElinux is disabled:
[root@www02 ~]# sestatus SELinux status: disabled [root@www02 ~]#
Any idea?? Could be a bug??
Host is a CentOS 6.5 x86_64 fully patched.
Thanks. _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
-
C. L. Martinez -
John Doe -
Marios Zindilis -
Steven Tardy