Nizamudeen A [Thu, 20 Feb 2025 09:10:28 +0000 (14:40 +0530)]
Merge pull request #61225 from rhcs-dashboard/rgw-ratelimit-integration1
mgr/dashboard: Rgw ratelimit feature for user and bucket
Reviewed-by: Aashish Sharma <aasharma@redhat.com> Reviewed-by: Ankush Behl <cloudbehl@gmail.com> Reviewed-by: Mark Nelson <mnelson@redhat.com> Reviewed-by: Nizamudeen A <nia@redhat.com> Reviewed-by: Naman Munet <nmunet@redhat.com>
mgr/dashboard: add actions to create, edit and delete smb join-auth and usersgroups resources
Add join-auth and usersgroups resources management and improve the way to select those from the smb cluster form using a dropdown.
Add option to navigate to join-auth/usersgroups resource create form from smb form
Add some additional fixes left over from previous work, such as adding helper texts or adding missing smb cluster form fields
Fixes: https://tracker.ceph.com/issues/69483 Signed-off-by: Pedro Gonzalez Gomez <pegonzal@redhat.com>
Ronen Friedman [Thu, 20 Feb 2025 08:01:44 +0000 (10:01 +0200)]
Merge pull request #61812 from ronen-fr/wip-rf-fstab
qa/standalone: do not use /etc/fstab as an always-there bytes source Reviewed-by: Patrick Donnelly <pdonnell@redhat.com> Reviewed-by: Ilya Dryomov <idryomov@gmail.com> Reviewed-by: Radoslaw Zarzynski <rzarzyns@redhat.com>
Max Kellermann [Mon, 17 Feb 2025 07:30:18 +0000 (08:30 +0100)]
include/unordered_{map,set}: remove, use `std` directly instead
This is only an alias for the `std` classes anyway. This effectively
reverts commit 8e8672047ccc which added these aliases, probably
because C++11 was new and not all compilers supported it properly, but
that was 12 years ago and we're onl C++20 meanwhile.
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Ronen Friedman [Thu, 13 Feb 2025 16:53:53 +0000 (10:53 -0600)]
qa/standalone: do not use /etc/fstab as an always-there bytes source
Multiple tests use /etc/fstab when a small data file is
required as input. After all, as some comments say:
# something that is always there
Alas - it's not always there. Not in containers.
Replacing with a newly-created temporary file filled with
random bytes.
For completeness - replacing similar references to
/etc/resolv.conf (as a source for random objects) in
the standalone tests, too.
See: https://github.com/llvm/llvm-project/commit/f46d1463b835560d90ad3ac02b63c771e4ebe566
```
common/interruptible_future.h:1182:16: error: a template argument list is expected after a name prefixed by the template keyword [-Wmissing-template-arg-list-after-template-kw]
1182 | ).template handle_interruption(std::move(efunc));
| ^
1 error generated.
```
Matan Breizman [Mon, 17 Feb 2025 10:00:34 +0000 (10:00 +0000)]
crimson/common/operation: fix decltype usage
```
ceph/src/crimson/common/operation.cc
In file included from /home/matan/ceph/src/crimson/common/operation.cc:4:
/home/matan/ceph/src/crimson/common/operation.h:312:31: error: no member named 'events' in 'AggregateBlockingEvent<T>'
312 | typename decltype(event.events)::iterator iter;
| ~~~~~ ^
1 error generated.
```
Zac Dover [Tue, 18 Feb 2025 13:44:15 +0000 (23:44 +1000)]
doc/mgr: edit grammar and formatting of rgw.rst
Improve the grammar and correct the formatting of the "Upgrading root ca
certificates" procedure that was added to the documentation in https://github.com/ceph/ceph/pull/61867
Naman Munet [Wed, 22 Jan 2025 10:59:20 +0000 (16:29 +0530)]
mgr/dashboard: Add confirmation textbox for resource name on delete action
Before:
=====
User was able to delete a single or multiple critical resources like ( images, snapshots, subvolumes, subvolume-groups, pools, hosts , OSDs, buckets, file system, services ) by just clicking on a checkbox.
After:
=====
User now has to type the resource name that they are deleting in the textbox on the delete modal, and then only they will be able to delete the critical resource.
Also from now onwards multiple selection for deletions of critical resources is not possible. Hence, user can delete only single resource at a time. On the other side, non-critical resources can be deleted in one go.
The BucketCacheFixtureInotify1.List2Inotify1 test was failing due to the
count being wrong. Fix the count, and flesh out the test case to test
the cache for deleting and adding. Also, move to per-test fixture, so
tests can't interfere with eachother due to inotify latency.
Fixes: https://tracker.ceph.com/issues/68336 Signed-off-by: Daniel Gryniewicz <dang@fprintf.net>
Ilya Dryomov [Mon, 17 Feb 2025 11:41:51 +0000 (12:41 +0100)]
test/rbd_mirror: flush watch/notify callbacks in TestImageReplayer
TestImageReplayer establishes its own (i.e. outside of the SUT code)
watch on the header of the remote image to be able to synchronize the
execution of the test with certain notifications. This watch is
established before the remote image is opened and is teared down until
after the remote image is closed but while the image replayer is still
running. The flush that is part of image close sequence thus isn't
guaranteed to cover all callbacks, especially for snapshot-based
mirroring where UnlinkPeerRequest spawned from Replayer::unlink_peer()
generates a notification on the remote image for each completed unlink.
Since TestImageReplayer further immediately deletes C_WatchCtx, pretty
much any test can segfault when C_WatchCtx::handle_notify() is invoked
by TestWatchNotify infrastructure. Because it's a virtual method, the
segfault often involves a completely bogus instruction pointer:
Ronen Friedman [Tue, 11 Feb 2025 12:52:38 +0000 (06:52 -0600)]
osd/scrub: clean-up scrub-store IDs generation
The Scrub Store IDs (used to identify OMAP entries used for storing
object-specific error data) are generated by attaching some prefix to
the text representation of virtual (invented) hobjects.
This commit cleans up the code by moving the name generation
process (creating temporary un-needed objects, then
extracting the to_str() representation of their identity)
into separate functions.
N Balachandran [Sat, 15 Feb 2025 13:26:31 +0000 (18:56 +0530)]
rbd-mirror: fix possible recursive lock of ImageReplayer::m_lock
If periodic status update (LambdaContext which is queued from
handle_update_mirror_image_replay_status()) races with shutdown and
ends up being the last in-flight operation that shutdown was pending
on, we attempt to recursively acquire m_lock in shut_down() because
m_in_flight_op_tracker.finish_op() is called with m_lock (and also
m_threads->timer_lock) held. These locks are needed only for the call
to schedule_update_mirror_image_replay_status() and should be unlocked
immediately.
Fixes: https://tracker.ceph.com/issues/69978 Co-authored-by: Ilya Dryomov <idryomov@gmail.com> Signed-off-by: N Balachandran <nithya.balachandran@ibm.com>
Adam King [Fri, 7 Feb 2025 21:08:30 +0000 (16:08 -0500)]
qa/cephadm: stop cephadm mgr module during cleanup
We recommend this generally when tearing down a cluster
and do so in cephadm-ansible. Doing this avoids any
potential timings where the cephadm mgr module attempts
to undo any daemon removals as we're cleaning up
projects / ceph.git / log