hi Nathan, On Thu, Nov 7, 2024 at 4:44 AM Nathan MALO <nathan.malo@gmail.com> wrote:

Hi,

I come to you with a quite simple question.

I am running a REEF cluster in 18.2.2, deployed in k3s via rook 1.14.8.

I want to enable per user and per bucket performance counters on my RGW.

According to this doc https://docs.ceph.com/en/latest/radosgw/metrics/#user-bucket-counter-caches I need to run ceph config set client.rgw rgw_{user,bucket}_counters_cache true.

you're looking at docs for the 'latest' development branch of ceph, rather than a specific release. this particular feature was added for squid: https://docs.ceph.com/en/squid/radosgw/metrics/#user-bucket-counter-caches and unfortunately does not exist for the reef release: https://docs.ceph.com/en/reef/radosgw/metrics/#user-bucket-counter-caches returns 404

But when I run the commands in the ceph toolbox, I end up with the following errors:

bash-4.4$ ceph config set client.rgw rgw_user_counters_cache true

Error EINVAL: unrecognized config option 'rgw_user_counters_cache'

bash-4.4$ ceph config set client.rgw rgw_bucket_counters_cache true

Error EINVAL: unrecognized config option 'rgw_bucket_counters_cache'

This doc https://www.ibm.com/docs/en/storage-ceph/7?topic=counters-viewing-performanc... mentions a different command regarding the client.rgw.

So I did the following:

bash-4.4$ ceph auth list | grep rgw

client.bootstrap-rgw

caps: [mon] allow profile bootstrap-rgw

client.rgw.ceph.s3.a

bash-4.4$ ceph config set client.rgw.ceph.s3.a rgw_user_counters_cache true

Error EINVAL: unrecognized config option 'rgw_user_counters_cache'

bash-4.4$ ceph config set client.rgw.ceph.s3.a rgw_bucket_counters_cache true

Error EINVAL: unrecognized config option 'rgw_bucket_counters_cache'

When I run all this commands with --force, the keys are created. I restarted my RGWs pods but the keys do not seem to be taken into account as I do not have any ceph_rgw_op* metric.

Did anyone successfully managed to enable this ?

What am I missing ?

Thank your for your help ! _______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io