Today Koji filled up its root partition which is designed to be small. In
part, there was 2.0 GB of httpd logs due to the API intensive use of Koji.
These logs are currently rotated and not compressed in anyway.
How should we handle this? A few ideas for discussion:
1) Modify logrotate configuration to try to compress the logs
2) Setup a cron job to compress the rotated logs
3) Rotate the logs less
1) We need to setup foreman-infra mailing list, we had one set by
Dominic but I am not sure we still have access to this. Greg, can you
set up something for us and add our infra team? Whatever this is, but
personally I'd prefer emails :)
I had an email about this on Saturday but I haven't read it until
today. I am bad in this.
2) By default httpd rotation has delayed compression so on busy day,
it can create one or two huge files until this is compressed. Although
we can tune logrotate, I've done this:
I moved and symlinked the following directories to local ephemeral volume:
Restarted httpd making sure there are no open jobs. Everything is fine
now, report issues.
···
On Mon, Dec 4, 2017 at 4:25 PM, Eric D Helms <ericdhelms@gmail.com> wrote:
Today Koji filled up its root partition which is designed to be small. In
part, there was 2.0 GB of httpd logs due to the API intensive use of Koji.
These logs are currently rotated and not compressed in anyway.
How should we handle this? A few ideas for discussion:
1) Modify logrotate configuration to try to compress the logs
2) Setup a cron job to compress the rotated logs
3) Rotate the logs less
We also noticed about 850MB of mrepo cache.
--
Eric D. Helms
Red Hat Engineering
--
You received this message because you are subscribed to the Google Groups
"foreman-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an
email to foreman-dev+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
