we need to maintain a fork of keepalived container image in order
to push it to quay.io so we can get around the docker rate limitation
Fixes: https://tracker.ceph.com/issues/54287
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
client: Inode::hold_caps_until is time from monotonic clock now.
Reviewed-by: Venky Shankar <vshankar@redhat.com>
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
mds: do not assert early on when issuing client leases
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
mds: fix crash when exporting unlinked dir
Reviewed-by: Venky Shankar <vshankar@redhat.com>
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Reviewed-by: Xiubo Li <xiubli@redhat.com>
this is done in order to better estimate the origin of the end-to-end
operation latency. and see how much of it is due to the objecter and how
much of it is due to the client that uses it (e.g. the RGW)
Signed-off-by: yuval Lifshitz <ylifshit@redhat.com>
osd/scrub: do not start scrubbing if the PG is snap-trimming
Reviewed-by: Samuel Just <sjust@redhat.com>
Reviewed-by: Laura Flores <lflores@redhat.com>
it can be convenient for automation tool if `orch status` output reports
whether the upgrade is currently paused.
Fixes: https://tracker.ceph.com/issues/55843
Signed-off-by: Guillaume Abrioux <gabrioux@redhat.com>
crimson/os/seastore/backref: don't merge in-cache backrefs that has already been released
Reviewed-by: Samuel Just <sjust@redhat.com>
Reviewed-by: Yingxin Cheng <yingxin.cheng@intel.com>
Now that RadorStore's get_new_req_id() just returns a random number, use
that as the default implementation for SAL.
Signed-off-by: Daniel Gryniewicz <dang@redhat.com>
Enable snapshot mirroring from the Pools -> Image
Also show the mirror-snapshot in the image where snapshot is enabled
When parsing images if an image has the snapshot mode enabled, it will
try to run commands that don't work with that mode. The solution was
not running those for now and appending the mode in the get call.
Fixes: https://tracker.ceph.com/issues/55648
Signed-off-by: Pere Diaz Bou <pdiazbou@redhat.com>
Signed-off-by: Nizamudeen A <nia@redhat.com>
Signed-off-by: Aashish Sharma <aasharma@redhat.com>
Signed-off-by: Avan Thakkar <athakkar@redhat.com>
codeowners: remove api team
Reviewed-by: Sarthak Gupta <sarthak.dev.0702@gmail.com>
Reviewed-by: nSedrickm <NOT@FOUND>
Reviewed-by: Nizamudeen A <nia@redhat.com>
Move the relevant ceph_assert()'s when the MDS is issuing leases
to clients since clients may not be capable of reading projected
linkages.
Fixes: http://tracker.ceph.com/issues/54701
Signed-off-by: Venky Shankar <vshankar@redhat.com>
* refs/pull/46516/head:
doc/dev/developer_guide/testing_integration_tests: document how to test custom kernels
Reviewed-by: Xiubo Li <xiubli@redhat.com>
Reviewed-by: Anthony D Atri <anthony.datri@gmail.com>
Reviewed-by: Venky Shankar <vshankar@redhat.com>
qa/cephfs: set omit_sudo False when sudo is set to True
Reviewed-by: Venky Shankar <vshankar@redhat.com>
Reviewed-by: Dhairya Parmar <dparmar@redhat.com>
Reviewed-by: Nikhilkumar Shelke <nshelke@redhat.com>
qa/cephfs: fix minor bug in caps_helper.py's run_mon_cap_tests()
Reviewed-by: Venky Shankar <vshankar@redhat.com>
Reviewed-by: Ramana Raja <rraja@redhat.com>
* extract the code to set logging fstream into a dedicated function
* do not reset logging until the end of the seastar application.
before this change, `reset_logger` is created in the
`if (auto log_file = local_conf()->log_file; !log_file.empty())` branch,
so its life cycle ends when the `if` block ends. in other words,
the cerr fstream is used for logging after the `if` block ends.
this is not the expected behavior.
after this changge, `reset_logger` is created out of the `if` block.
so we won't reset the logger back to `cerr` until the lambda passed to
`seastar::async()` exits.
Signed-off-by: Kefu Chai <tchaikov@gmail.com>
