I've been having some trouble with logrotate, it seemingly does not do what I'm telling it to do.
Environment:
- Centos 6.4
- logrotate 3.7.8
My /etc/logrotate.conf file has the following:
# rotate log files weekly
daily
# keep 4 weeks worth of backlogs
rotate 30
# create new (empty) log files after rotating old ones
create
# use date as a suffix of the rotated file
# dateext
# uncomment this if you want your log files compressed
compress
compresscmd /usr/bin/bzip2
uncompresscmd /usr/bin/bunzip2
compressext .bz2
# RPM packages drop log rotation information into this directory
include /etc/logrotate.d
And the logrotate file I'm having trouble with is this (for elasticsearch, located at /etc/logrotate.d/elasticsearch):
/var/log/elasticsearch/*.log {
missingok
notifempty
copytruncate
postrotate
rm -rf /var/log/elasticsearch/*.log.$(date +%Y)*
size 1k
rotate 7
daily
}
First of all, it's not respecting my rotate 7 configuration, when I run logrotate -d /etc/logrotate.conf, I get a line saying:
rotating log /var/log/elasticsearch/gravity-es-prod02.log, log->rotateCount is 30
... and a bunch of statements saying it's rotating 30 different *.bz2 files.
Secondly, I keep ending up with a logfile that is named gravity-es-prod02.log.2015-12-01 (and any previous dates since my last manual cleanup), despite not having dateext enabled. These also don't get cleaned up, so I added the postrotate line to manually clean it up, but apparently that doesn't work correctly either.
EDIT The cron file that runs the logrotate script is pretty standard:
#> cat /etc/cron.daily/logrotate
#!/bin/sh
/usr/sbin/logrotate /etc/logrotate.conf >/dev/null 2>&1
EXITVALUE=$?
if [ $EXITVALUE != 0 ]; then
/usr/bin/logger -t logrotate "ALERT exited abnormally with [$EXITVALUE]"
fi
exit 0