wait_for_complete tests currently hanging with crimson-osd
and are temporary disabled in Crimson suite.
Signed-off-by: Matan Breizman <mbreizma@redhat.com>
*Adding fs_name as a field option in perf_stats o/p
*perf stats command incorrect output with non-existing mds_rank filter
If `ceph fs perf stats` runs with non-existing mds_rank filter,
even then it shows all the clients `client_metadata` and `global_metrics`
Fixes: https://tracker.ceph.com/issues/56162
Fixes: https://tracker.ceph.com/issues/56169
Signed-off-by: Neeraj Pratap Singh <neesingh@redhat.com>
in some corner cases on single key erase sence,
for e.g. we have three nodes which are brothers:
1 2 3 4 5 | 6 7 8 | 9 10 11, node min size is 3.
When erasing 11 will trigger balance instead of
merge, the middle nodes will adjust to 6 7, this
will not ensure the all nodes above min size.
This corner case might not be harmful but if we
want to erase 10,11 at one time (range remove wip
need this), this will trigger balance in past
policy which will make both two nodes below min
size and we might need at least one more merge.
This is actually because we cannot ensure a node
to merge in tree is above min size.
Signed-off-by: Xinyu Huang <xinyu.huang@intel.com>
doc/cephfs: note regarding start time time zone
Reviewed-by: Venky Shankar <vshankar@redhat.com>
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Reviewed-by: Anthony D'Atri <anthony.datri@gmail.com>
Reviewed-by: Dhairya Parmar <dparmar@redhat.com>
Adapt the XML ListBucket response to contain well-formed JSON,
matching the implied schema of awscli:
$ aws --endpoint-url http://localhost:8000 s3api list-objects --bucket testbucket
suggested by Casey in review.
Fixes: https://tracker.ceph.com/issues/53455
Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>
In general, e.g., in the S3 ListBucket response, the
output Formatter is already of the type requested in the HTTP
Accept header.
Therefore, while there are still many instances where the of
Formatter is assumed to be XML, it appears necessarily safe
and seemingly correct to extend this assumption to end_header(...),
which sends the response content-type.
Fixes: https://tracker.ceph.com/issues/55680
Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>
This is a pure cleanup. The method to print an RGWFormat object
as a MIME type is now called to_mime_type().
Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>
Some code was using old, non-zipper versions of code that existed in
zipper. Use the zipper versions instead.
Signed-off-by: Daniel Gryniewicz <dang@redhat.com>
- RadosStore had a dummy implementation of Zone that just passed through
to the Zone service to get info about the local zone. Implement the
Zone API for real, so that we can reference multiple zones.
- Convert get_id() to string; it was only ever used as a string anyway.
- Add get_zone_by*() to zonegroup and use it apporpriately
Signed-off-by: Daniel Gryniewicz <dang@redhat.com>
The get_zonegroup() that looks up by ID doesn't belong in the zone. It
was there because of a conflation of zone as an object an zone as an
interface. Move it to Store where it belongs.
Signed-off-by: Daniel Gryniewicz <dang@redhat.com>
Added periodic calls to all PGs for which the OSD is the primary,
asking for a scrub statistics update.
This allows operator queries (e.g. 'pg dump pgs') to present
up-to-date scrub duration, "scrub is blocked" duration, etc'.
Signed-off-by: Ronen Friedman <rfriedma@redhat.com>
mgr/cephadm: mock CephadmHttpServer run in unit tests
Reviewed-by: Michael Fritch <mfritch@suse.com>
Reviewed-by: Redouane Kachach <rkachach@redhat.com>
This patch adds the capability to remove the stale snapshot user
metadata while loading the subvolume if it is present. It can't
be done in 'SubvolumeBase.discover' since v1 and v2 snapshot paths
are different. This is done just after the discover before returning
the specific version object.
Fixes: https://tracker.ceph.com/issues/55976
Signed-off-by: Kotresh HR <khiremat@redhat.com>
When the osd is full, if the snapshot has metadata set, it
can't be removed as user metadata can't be removed when osd
is full. This patch provides a way to remove the snapshot
with 'force' option while keeping the corresponding metadata
which gets removed on subvolume discover when it finds space.
Fixes: https://tracker.ceph.com/issues/55976
Signed-off-by: Kotresh HR <khiremat@redhat.com>
The 'metadata_mgr.flush()' used to truncate the config file
before flushing the new config data. This could lead to an
empty config file when there is no space to write new config
data. This patch handles this scenario by writing it to
temporary file and rename it to config file. This would
retain the config file without truncating it.
Also, there are bunch of places which wasn't handling
'MetadataMgrException' because of this. Fixed those.
Fixes: https://tracker.ceph.com/issues/55976
Signed-off-by: Kotresh HR <khiremat@redhat.com>
As an example this PR displays the list of Ceph Auth users, keys and
caps.
It tries to minimize the amount of UI code required by feeding a generic
table-like component (RestTable) with backend-generated JSON data.
This is just a proof of concept and there's lot of room for improvement.
Fixes: https://tracker.ceph.com/issues/52701
Signed-off-by: Ernesto Puerta <epuertat@redhat.com>
Signed-off-by: Pere Diaz Bou <pdiazbou@redhat.com>
Signed-off-by: Nizamudeen A <nia@redhat.com>
crimson/os/seastore/journal: move bufferlists to reduce the amount of memory copies
Reviewed-by: Chunmei Liu <chunmei.liu@intel.com>
Reviewed-by: Yingxin Cheng <yingxin.cheng@intel.com>
