2024-07-10T12:00:47.205Z

<Kaleb> wrt QAT, ceph-19 cmake (CMakeLists.txt) seems to assume that the whole world is either x86_64 or aarch64 and only bothers to disable QAT on aarch64.  I submit that QAT should be disabled by default, and (only) enabled on x86_64. That way it will do the right thing on ppc64le and s390x, amongst others.  (Think Fedora, CentOS, Ubuntu, etc.)

2024-07-10T12:53:53.637Z

<Casey Bodley> Ken recently changed this in <https://github.com/ceph/ceph/pull/57479>

2024-07-10T12:55:07.037Z

<Casey Bodley> it doesn't look like it's in 19.1.0 but should be in 19.1.1

2024-07-10T12:56:27.418Z

<Kaleb> right (see #ann_arbor)   Does his change not disable QAT on x86_64 and amd64? That's what it looks like to me. Maybe I'm having a senior moment?

2024-07-10T12:57:46.295Z

<Kaleb> right (see my comment/question in #ann_arbor)   Does his change not disable QAT on x86_64 and amd64? That's what it looks like to me. Maybe I'm having a senior moment?

2024-07-10T13:00:14.295Z

<Casey Bodley> the syntax for CMAKE_DEPENDENT_OPTION is confusing. i did check the builds in <https://github.com/ceph/ceph/pull/57479#pullrequestreview-2058199127> to verify that it was working for x86_64

2024-07-10T13:33:48.014Z

<Rost Khudov> Hello, I have a question regarding multisite test suite for rgw
I am following documentation: <https://github.com/ceph/ceph/blob/main/src/test/rgw/test_multi.md>
When running <https://github.com/ceph/ceph/blob/main/src/test/rgw/test_multi.py> it uses `mstart.sh` to start cluster, but does it call `mstop.sh` when tests are finished? when tests are failed?

2024-07-10T13:34:36.940Z

<Casey Bodley> no, it leaves the cluster intact

2024-07-10T13:35:30.917Z

<Casey Bodley> you can rerun tests against that cluster by adding `no_bootstrap = 1` to your config. otherwise you have to mstop manually

2024-07-10T13:36:50.706Z

<Rost Khudov> but how can I mstop manually, if I don't know the name of the cluster which is created when `test_multi,py` is executed?

2024-07-10T13:41:30.873Z

<Rost Khudov> Or cluster name is always c1?
<https://github.com/ceph/ceph/blob/main/src/test/rgw/test_multi.py#L239>

2024-07-10T13:44:15.393Z

<Casey Bodley> yeah, c1 through cN depending on number of zones configured

2024-07-10T13:45:13.261Z

<Rost Khudov> so I have to execute following:
```mstop.sh c1
mstop.sh c2
...
mstop.sh cN```
is it correct?

2024-07-10T13:45:43.521Z

<Casey Bodley> right

2024-07-10T13:45:54.551Z

<Rost Khudov> okay, thank you

2024-07-10T13:57:41.359Z

<Casey Bodley> is there something specific you're trying to test or work on? there's a weekly rgw meeting in ~90 minutes if you'd like to meet some of the team and discuss (see 'RGW refactoring' on <https://ceph.io/en/community/meetups/#community>)

2024-07-10T14:00:20.140Z

<Rost Khudov> Just trying to make it work and had a feeling that after finishing of multisite, cluster wasn't stopped
but nothing was mentioned in the doc either. But for example for [RabbitMQ notification tests](https://bbgithub.dev.bloomberg.com/ceph/ceph/tree/6f5a6f4773c5d6773ebdb5b3a24ce12e520f5019/src/test/rgw/bucket_notification#rabbitmq-tests) I did see a documentation saying how to stop vstart cluster

2024-07-10T14:01:12.398Z

<Rost Khudov> Just trying to make it work and had a feeling that after finishing of multisite, cluster wasn't stopped
but nothing was mentioned in the doc either. But for example for [RabbitMQ notification tests](https://github.com/ceph/ceph/blob/main/src/test/rgw/bucket_notification/README.rst#rabbitmq-tests) I did see a documentation saying how to stop vstart cluster

2024-07-10T14:52:29.511Z

<Kaleb> okay

2024-07-10T15:26:46.911Z

<Casey Bodley> weekly rgw meeting starting soon in [ <https://pad.ceph.com/p/rgw-weekly](https://meet.google.com/oky-gdaz-ror> )