Hi Casey, If you are using the bi-directional replication. When ever you modify the quota on Zone A. It will modify the master Zone as well. If you one use way replication. I.e that is from the master zone to zone A, it will not Regards Natanam Sethu On Thu, 6 Aug 2026 at 18:49, Casey Bodley via ceph-users <ceph-users@ceph.io> wrote:

On Thu, Aug 6, 2026 at 3:44 AM Danila Reznichuk via ceph-users <ceph-users@ceph.io> wrote:

...

Hi,

I have a question about account quotas in an RGW multisite setup.

My configuration is fairly simple:

Main zonegroup with the master zone. Archive zonegroup with two zones: A and B.

I created an account on the master zone of the main zonegroup and set the account quota (--quota-scope account) to 1 GiB. As expected, this change was synchronized to all other zones in all zonegroups.

Then, by mistake, I changed the quota for the same account on zone A of the archive zonegroup to 2 GiB. This change was not synchronized to the other zones. That seems reasonable to me and could even be considered a feature, since different zonegroups may represent different locations with different storage policies or usage patterns.

After some testing, I found that updating the account quota again on the master zone of the main zonegroup overwrites the locally modified quota on zone A.

So I have a few questions:

Is this the intended behavior? Is there a supported way to configure different account quotas for different zones or zonegroups? If not, is the account quota expected to be globally consistent across the entire multisite deployment?

the intent of multisite is to keep all metadata (accounts, users, buckets, etc) consistent across zones, with the metadata master zone as the source of truth. there are several radosgw-admin commands that can modify metadata on secondary zones, but they're supposed to print a warning message and require --yes-i-really-mean-it. that logic was missing from the quota commands, so i've raised https://tracker.ceph.com/issues/79120 to track the fix and its backports

...

Thanks! _______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io

_______________________________________________ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-leave@ceph.io