The mixed type tests will change strategy after the idempotency test so
we need to handle that in test playbook.
Signed-off-by: Andrew Schoen <aschoen@redhat.com>
If we fail to load a ceph configureation file when trying to get the
block.db size then just use defaults instead of throwing an error.
Signed-off-by: Andrew Schoen <aschoen@redhat.com>
This will pick a strategy, filter out any devices already been used by
ceph and then pick a strategy again. If the strategy has changed the
call should error, if the strategy is the same proceed. If there are no
unused devices then the command is a noop.
Signed-off-by: Andrew Schoen <aschoen@redhat.com>
This will indicate if the command would result in any OSDs being created
or not. Other tooling can use that key for idempotency checks.
Signed-off-by: Andrew Schoen <aschoen@redhat.com>
cmake: do not use plain target_link_libraries(rgw_a ...)
Reviewed-by: Laura Paduano <lpaduano@suse.com>
Reviewed-by: Kai Wagner <kwagner@suse.com>
Reviewed-by: Casey Bodley <cbodley@redhat.com>
this addresses following error:
CMake Error at src/rgw/CMakeLists.txt:178 (target_link_libraries):
The plain signature for target_link_libraries has already been used
with
the target "rgw_a". All uses of target_link_libraries with a target
must
be either all-keyword or all-plain.
The uses of the plain signature are here:
* src/rgw/CMakeLists.txt:168 (target_link_libraries)
Signed-off-by: Kefu Chai <kchai@redhat.com>
Make sure the mbuf and ethdev libraries are created as they export
required symbols.
Fixes: http://tracker.ceph.com/issues/36341
Signed-off-by: Brad Hubbard <bhubbard@redhat.com>
The dashboard API now can delete all compression options if the
compression mode is set to 'unset' during edit. Other dismissed options
will be overwritten with the reset value on submit, which will cause
the deletion of the previously set settings.
Fixes: https://tracker.ceph.com/issues/36355
Signed-off-by: Stephan Müller <smueller@suse.com>
It is now commented out like it was before,
but I've added a comment what happened during this test with the QA
system. The problem was that even with only a increase of 1 PG the QA
cluster went into a cluster warning state and did not recover in time.
The QA coverage timeout is 2 minutes.
I could not reproduce this behavior with a local cluster, but I've
added a loop to wait until pgp and pg number are equal and the cluster
is in a healthy state again. This can take locally about 5 seconds.
The internal loop has a timeout of 3 minutes.
Fixes: https://tracker.ceph.com/issues/36362
Signed-off-by: Stephan Müller <smueller@suse.com>
The dashboard backend can now unset all set compression arguments if the
compression mode is switched to 'unset'. In the case of 'unset' Ceph
itself will only delete the 'compression_mode' argument, not all other
set arguments. The other arguments that should be removed, too, are
added to the update arguments in order to delete all set arguments.
Fixes: https://tracker.ceph.com/issues/36355
Signed-off-by: Stephan Müller <smueller@suse.com>
