in this change, along with LockPolicy, src/common/lock_* are completely
removed.
instead of using LockPolicy based template specialization of LockMutex,
etc, it would be simpler if we can just rely on WITH_SEASTAR
preprocessor macro to tell if we are compiling code for crimson or not.
but please bear in mind, we cannot link against the plain libceph-common
in crimson anymore.
Signed-off-by: Kefu Chai <kchai@redhat.com>
for the sake of completeness, also pave the road to ditch
SharedMutex<LockPolicy>. please note, in code used by crimson, we are
not using shared_mutex or SharedMutex yet. so we are not adding
the dummy shared_mutex for WITH_SEASTAR.
Signed-off-by: Kefu Chai <kchai@redhat.com>
When connection stuck into odd state, we need to let connection timeout.
If send_message could update last_active, it won't play the role as
expected.
So we move last_active updated to the place actually should.
Signed-off-by: Haomai Wang <haomai@xsky.com>
Signed-off-by: Kefu Chai <kchai@redhat.com>
Introducing Clay codes to ceph. These codes are optimal
in terms of network, disk traffic utilized during repair of a lost chunk.
This feature also provides gradual increase in network, disk traffic with
respect to the number of lost chunks.
Authors: Myna, Elita.
Fixes: http://tracker.ceph.com/issues/19278
Signed-off-by: Myna V <mynaramana@gmail.com>
build: mgr: check for python's ssl version linkage
Reviewed-by: Tim Serong <tserong@suse.com>
Reviewed-by: John Spray <john.spray@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>
tools: backport-create-issue: rate-limit to avoid seeming like a spammer
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Reviewed-by: Abhishek Lekshmanan <abhishek@suse.com>
The only classes in rgw_sync.cc that remain
with ldouts are:
- RGWBackoffControlCR
- RGWShardCollectCR
- RGWAsyncMetaStoreEntry
- RGWAsyncMetaRemoveEntry
- PurgePeriodLogsCR
- take_min_status
- MetaMasterTrimShardCollectCR
- MetaMasterStatusCollectCR
- MetaMasterTrimCR
- MetaPeerTrimCR
- MetaTrimPollCR
These classes do not have RGWMetaSyncEnv members.
Signed-off-by: Ali Maredia <amaredia@redhat.com>
rgw: apply quota config to users created via external auth
Reviewed-by: Matt Benjamin <mbenjami@redhat.com>
Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@gmail.com>
The IO was being unblocked before object map was created, allowing
a potential copyup request to fail to update a still-to-be-created
object map.
Fixes: http://tracker.ceph.com/issues/24516
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
The copyup state machine needs to iterate over all object maps to update
the existence for the object. If an snapshot is being removed concurrently,
it's possible to invalidate the object map for the image.
Fixes: http://tracker.ceph.com/issues/24516
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
as the latest SPDK deprecated and removed spdk_pci_get_device(),
we cannot use the PCI device's serial number as its identifier and the
filename of the file representing the device. in this change, we
are using the PCI device's selector instead, it is also used by SPDK as
part of the transport id.
Tested-by: Tone Zhang <tone.zhang@arm.com>
Signed-off-by: Kefu Chai <kchai@redhat.com>
Clang complains:
```
In file included from /home/jenkins/workspace/ceph-master/src/osd/ReplicatedBackend.cc:15:
In file included from /home/jenkins/workspace/ceph-master/src/osd/ReplicatedBackend.h:18:
In file included from /home/jenkins/workspace/ceph-master/src/osd/PGBackend.h:27:
In file included from /home/jenkins/workspace/ceph-master/src/osd/PGTransaction.h:25:
/home/jenkins/workspace/ceph-master/src/common/inline_variant.h:88:64: error: expected class name
struct generic_visitor : boost::static_visitor<Result>, boost::noncopyable
^
1 error generated.
gmake[2]: *** [src/osd/CMakeFiles/osd.dir/build.make:141: src/osd/CMakeFiles/osd.dir/ReplicatedBackend.cc.o] Error 1
gmake[2]: *** Waiting for unfinished jobs....
```
Signed-off-by: Willem Jan Withagen <wjw@digiware.nl>
Sporadically the rbd-mirror fsx stress test would fail due to very
slow sync times due to overloaded clusters. Attempt to wait for all
images to be replicated before proceeding with the comparison.
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
