Hi Eugen,
Yes i have played arround with some of them, with the most obvious ones.
They are all false by default:
:~# ceph-conf -D | grep syslog
clog_to_syslog = false
clog_to_syslog_facility = default=daemon audit=local0
clog_to_syslog_level = info
err_to_syslog = false
log_to_syslog = false
mon_cluster_log_to_syslog = default=false
mon_cluster_log_to_syslog_facility = daemon
mon_cluster_log_to_syslog_level = info
Loggically, if you turn off ops logs and it applies instantly I assume
these
configs from above are respected as intended. Still, I have turned on and
off log_to_syslog to no avail.
Interesting is the ops log file on disk is opened by the radosgw process in
the container.
So, inside a container, radosgw services ignores the rgw_ops_log_rados when
true.
Other configs like rgw_max_concurrent_requests and rgw_enable_ops_log are
working as expected.
--
Paul
On Thu, Nov 28, 2024 at 10:30 AM Eugen Block <eblock@xxxxxx> wrote:
> I haven't played with rgw_ops yet, but have you looked at the various
> syslog config options?
>
> ceph config ls | grep syslog
> log_to_syslog
> err_to_syslog
> clog_to_syslog
> clog_to_syslog_level
> clog_to_syslog_facility
> mon_cluster_log_to_syslog
> mon_cluster_log_to_syslog_level
> mon_cluster_log_to_syslog_facility
>
> Maybe one of them is what you're looking for.
>
> Zitat von Paul JURCO <paul.jurco@xxxxxxxxx>:
>
> > Hi!
> > Currently I have limitted the optput of rgw log to syslog from rsyslog
> (as
> > suggested by
> > Anthony), limitted docker logs from daemon.json.
> > I still get ops logs written to both logs pool and ops log file
> > (ops-log-ceph-client.rgw.hostname.log).
> >
> > How to stop logging ops log on rgw disk and keep logs on logs pool?
> > Current config:
> > global advanced debug_rgw
> > 0/0
> > global advanced
> > rgw_enable_ops_log true
> > global advanced
> > rgw_ops_log_rados true
> >
> > Thank you!
> >
> > --
> > Paul Jurco
> >
> >
> > On Fri, Nov 22, 2024 at 6:11 PM Paul JURCO <paul.jurco@xxxxxxxxx> wrote:
> >
> >> Hi,
> >> we recently migrated to cephadm from ceph-deploy a 18.2.2 ceph cluster
> >> (Ubuntu with docker).
> >> RGWs are separate vms.
> >> We noticed syslog increased a lot due to rgw's access logs sent to it.
> >> And because we use to log ops, a huge ops log file on
> >> /var/log/ceph/cluster-id/ops-log-ceph-client.rgw.hostname-here.log.
> >>
> >> While having "rgw_ops_log_rados": "true", oplogs goes to both file and
> >> rados pool for logs.
> >> If false it doesn't log anything, as expected.
> >> How to stop dockered rgws to log to syslog and to a file on disk, but to
> >> keep opslog in logs pool?
> >>
> >> Config is:
> >> global basic
> >> log_to_journald false
> >> global advanced
> >> rgw_enable_ops_log false
> >> global advanced
> >> rgw_ops_log_rados true
> >>
> >> A few hours later after after enabling it back, after massive cleanup,
> it
> >> does logging ops but only to files.
> >> How to get ops logs in rados pool and access log to a file on disk but
> not
> >> on syslog?
> >> I have add this to daemon.json to limit access logs to accumulate on
> >> /var/log/docker/containers/rand/rand/json.log file:
> >>
> >> {
> >> "log-driver": "local",
> >> "log-opts": {
> >> "max-size": "512m",
> >> "max-file": "3"
> >> }
> >> }
> >>
> >>
> >> Thank you!
> >> Paul
> >>
> > _______________________________________________
> > ceph-users mailing list -- ceph-users@xxxxxxx
> > To unsubscribe send an email to ceph-users-leave@xxxxxxx
>
>
> _______________________________________________
> ceph-users mailing list -- ceph-users@xxxxxxx
> To unsubscribe send an email to ceph-users-leave@xxxxxxx
>
_______________________________________________
ceph-users mailing list -- ceph-users@xxxxxxx
To unsubscribe send an email to ceph-users-leave@xxxxxxx
