Ceph

10/19/2022

12:04 PM bluestore Bug #55324: rocksdb omap iterators become extremely slow in the presence of large delete range to...

> I see this was backported in: https://github.com/ceph/ceph/pull/45963 but was later reverted in https://github.com/...

06/16/2022

12:40 PM RADOS Backport #53338: pacific: osd/scrub: src/osd/scrub_machine.cc: 55: FAILED ceph_assert(state_cast<...

It looks like this backport has been merged in https://github.com/ceph/ceph/pull/45374, and released in 16.2.8, so I ...

05/16/2022

08:42 AM rgw Bug #55546: rgw: trigger dynamic reshard on index entry count rather than object count

I just want to note that if you want to scale `rgw_max_objs_per_shard` for versioned buckets, there's a special case ...

05/02/2022

07:13 AM rgw Backport #55499: quincy: Segfault when Open Policy Agent authorization is enabled

please link this Backport tracker issue with GitHub PR https://github.com/ceph/ceph/pull/46107
ceph-backport.sh versi...

07:12 AM rgw Backport #55500: pacific: Segfault when Open Policy Agent authorization is enabled

please link this Backport tracker issue with GitHub PR https://github.com/ceph/ceph/pull/46106
ceph-backport.sh versi...

04/19/2022

07:00 AM rgw Bug #52721: Bucket index entry not always created

J. Eric Ivancich wrote:
> This sounds like it could result from the bug fixed here:
>
> https://tracker.ceph.com/...

04/12/2022

10:50 AM rgw Bug #55286: Segfault when Open Policy Agent authorization is enabled

Pull request: https://github.com/ceph/ceph/pull/45873.

10:48 AM Ceph Revision ac71916c (ceph): rgw: Avoid segfault when OPA authz is enabled

When `rgw_use_opa_authz=true`, radosgw would segfault on any request that
didn't target a specific object or bucket, ...

10:47 AM rgw Bug #55286 (Pending Backport): Segfault when Open Policy Agent authorization is enabled

On Pacific, radosgw segfault when `rgw_use_opa_authz=true`:
```
2022-04-05T09:03:01.940+0000 7fa29c225700 -1 *** ...

03/29/2022

06:14 AM rgw Bug #52721: Bucket index entry not always created

We ran into the exact same issue, with a very similar setup: Ceph Pacific (16.2.7), 3 radosgw instances behind HAprox...