Use logrotate instead of custom script

[andsens]

Why don't you use logrotate for rolling over logs?
It's directly built into debian and supports restarting services and loads of other stuff.
You don't even need a cronjob, single file in /etc/logrotate.d and you're done.
It has loads of config options.

[marcuswhybrow]

Looks interesting. I am looking for something to provide purging of all logs, as well as world and server backups. This might be a partial or total solution.

[marcuswhybrow]

From the man page it looks as though it would be useful for rotating server.log, but not purging old backups.

Looks like it would be a better tool than the wheel reinvention I have in place at the moment.

[andsens]

rotate count
             Log  files  are  rotated  <count>  times before being removed or mailed to the address specified in a mail
             directive. If count is 0, old versions are removed rather then rotated.

[morbeo]

I've played around with a config file which you can put in /etc/logrotate.d/ . You can test configs with logrotate -d -f

# Sample wildcard
/opt/msm/servers/*/server.log {
# Running with cron.daily time
daily 
olddir /opt/msm/archives/logs/
# Keeps a fortnight
rotate 14 
# Don't rotate empty logs
notifempty 
}

The issue with this is old logs are located in separate folders for each server in /opt/msm/archive/logs/ which cannot be interpreted to the olddir directive. If a unified folder is used for all the servers there is a conflict of names, because all the logs are named server.log. What I think could work is something like this where each server is appended after its created with MSM

# Main config file to which each server is appended
daily
rotate 14
notifempty

Sample append in server_create() for example

echo -e "$SETTINGS_SERVER_STORAGE_PATH/$1/server.log { \n olddir $LOG_ARCHIVE_PATH/$1/ \n } " >> /etc/logrotate.d/msm

This will most likely work, but will be difficult to manage if a server is removed