EBLOCKLISTED has a very special meaning but happens to be an alias for
ESHUTDOWN. If the client gets blocklisted, we always want to propagate
EBLOCKLISTED error code since it's generated by the OSD.
For ManagedLock use case of indicating that an operation on the lock
raced with lock shut down, meaning that a higher level request can just
be restarted, ERESTART should do.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
needs_exclusive_lock() calls acquire_lock() with owner_lock held.
If lock acquisiton races with lock shut down, ManagedLock completes
ImageDispatch context directly and dispatch is retried immediately on
the same thread (due to DISPATCH_RESULT_RESTART). This results in
recursion into needs_exclusive_lock() and, barring locking issues, can
lead to unbounded stack growth if lock shut down takes its time.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Eliminate a race where a client is able to submit an operation after
WatchCtx2::handle_error() is invoked on its watch due to blocklisting.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
qa/upgrade: consistently use the tip of the branch as the start version
Reviewed-by: Ilya Dryomov <idryomov@redhat.com>
Reviewed-by: Josh Durgin <jdurgin@redhat.com>
Reviewed-by: Neha Ojha <nojha@redhat.com>
Reviewed-by: Laura Flores <lflores@redhat.com>
Also corrected line to use image: quay.ceph.io/ceph-ci/ceph:*
In order to make sure that latest versions of ceph is used as initial install
Fixes: https://tracker.ceph.com/issues/58914
Signed-off-by: Yuri Weinstein <yweinste@redhat.com>
When the zones replicate, allow disabling replication
for specific buckets using sync policy.
These are the semantics to be followed while resolving the policy
conflicts -
==================================================
zonegroup bucket Result
==================================================
enabled enabled enabled
allowed enabled
forbidden disabled
allowed enabled enabled
allowed disabled
forbidden disabled
forbidden enabled disabled
allowed disabled
forbidden disabled
In case multiple group policies are set to reflect for any sync pair
(<source-zone,source-bucket>, <dest-zone,dest-bucket>), the following
rules are applied in the order -
1) Even if one policy status is FORBIDDEN, the sync will be disabled
2) Atleast one policy should be ENABLED for the sync to be allowed.
Various cases tested are outlined here -
https://docs.google.com/document/d/19oBQA-bYxLBR4BnekA2DTwJJaTFvjAfrqAk9G3RGU0I/edit#heading=h.4qac9dpc76m
Signed-off-by: Soumya Koduri <skoduri@redhat.com>
Modernize prompts in example commands in file
fs-volume.rst and use a root prompt correctly instead
of using a non-privileged prompt when privileges
are needed.
Indent all JSON in example command outputs with
four spaces consistently instead of a mix of four and
two spaces.
Signed-off-by: Ville Ojamo <14869000+bluikko@users.noreply.github.com>
rgw: set init_check_compat when bucket sync status doesn't exist
Reviewed-by: Shilpa Jagannath <smanjara@redhat.com>
Reviewed-by: Mark Kogan <mkogan@redhat.com>
rgw: fix CopyObj crash after admin override
Reviewed-by: Matt Benjamin <mbenjamin@redhat.com>
Reviewed-by: Daniel Gryniewicz <dang@redhat.com>
Reviewed-by: Mark Kogan <mkogan@redhat.com>
The "append to the passed list" behavior is confusing and not what the
corresponding C API (rbd_watchers_list) or other similar C++ APIs (e.g.
list_lockers) do.
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
The Windows CI job started to fail as some libcephfs tests that use
ALLPERMS have been moved [1] to a separate file which doesn't have
the ALLPERMS definition.
/ceph/src/test/libcephfs/suidsgid.cc:240:36: error: ‘ALLPERMS’ was
not declared in this scope
240 | ASSERT_EQ(stx.stx_mode & (mode_t)ALLPERMS, before_mode);
We'll move this definition to compat.h so that we won't have to
redefine it in each file that uses it.
Note that we're moving the Windows "fs_compat.h" include up,
ensuring that the constants used by ALLPERMS are defined.
[1] 254eeb2730
Signed-off-by: Lucian Petrut <lpetrut@cloudbasesolutions.com>
When a metadata osd fails, an unregister_watch() error may occur,
resulting in an rgw cache invalidation.
By adding an unregister_done flag and when a register_watch() error ,
performing a reinit() operation again,
After the first reinit() failure, the register_watch() will be performed again
Fixes: https://tracker.ceph.com/issues/59217
Signed-off-by: lichaochao <lichaochao2_yewu@cmss.chinamobile.com>
