#8375 reeduce or redirect new mirrorlist logging
Closed: Fixed 4 years ago by kevin. Opened 4 years ago by kevin.

The new mirrorlist containers are logging a lot to syslog.

Basically for every hit:

Nov 11 14:30:05 proxy11.vpn.fedoraproject.org podman[45077]: mirrorlist: x.y.z.a found its best mirror from netblocks
Nov 11 14:30:05 proxy11.vpn.fedoraproject.org podman[45077]: INFO  mirrorlist_server > Found 60 possible mirrors
Nov 11 14:30:05 proxy11.vpn.fedoraproject.org podman[45341]: INFO  mirrorlist_server > mirrors_found after netblock 1
Nov 11 14:30:05 proxy11.vpn.fedoraproject.org podman[45341]: INFO  mirrorlist_server > mirrors_found after Internet2 1
Nov 11 14:30:05 proxy11.vpn.fedoraproject.org podman[45341]: INFO  mirrorlist_server > mirrors_found after country 1
Nov 11 14:30:05 proxy11.vpn.fedoraproject.org podman[45341]: INFO  mirrorlist_server > mirrors_found after geoip country 1

multiplied by all proxies and every request... it's a bit much for syslog. :)

Can we redirect it to the local mirrorlist log? or supress it entirely ?


Everything starting with INFO can be easily disabled by an environment variable (removing -e RUST_LOG=info from the mirrorlist systemd file).

We used to have mirrorlist: x.y.z.a found its best mirror from netblocks in the old implementation, but nothing actually uses it as far as I know. So I can also remove that.

On the next proxy playbook run, the container systemd files will have the INFO logging removed. I will also remove the mirrorlist: x.y.z.a found its best mirror from netblocks from the next mirrorlist-server release.

Hi adrian,

I was going to say could this info be moved to debug. It was something I was using for debug a long time ago but I never found a way to use this as a general information. However removing also works.

The new version will be completely silent and the current output will only be available when setting the logging variable to INFO.

Verified in staging that the containers are not logging any more. All existing statistics are using other logs to collect the data, so this log was really only for debugging purposes.

Now it has to be applied in production. Will make the ansible change tomorrow.

@smooge pushed this out and I can confirm the debugging is not there. Thanks!

Metadata Update from @kevin:
- Issue close_status updated to: Fixed
- Issue status updated to: Closed (was: Open)

4 years ago

Login to comment on this ticket.

Metadata