os/bluestore: call fault_range prior to looking for blob to reuse
Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>
os/ceph-bluestore-tool: bluefs-bdev-expand asserts if no WAL
Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>
Reviewed-by: Sage Weil <sage@redhat.com>
mgr/dashboard: Enable read only users to read again
Reviewed-by: Ernesto Puerta <epuertat@redhat.com>
Reviewed-by: Laura Paduano <lpaduano@suse.com>
Reviewed-by: Tatjana Dehler <tdehler@suse.com>
EPEL7 has switched over to python3.6 as the main python3. and we started
packaging python bindings for python3.6 since
https://github.com/ceph/ceph-build/pull/1283
Signed-off-by: Kefu Chai <kchai@redhat.com>
'ceph ... > /etc/ceph/ceph.conf' fails because bash truncates ceph.conf
before the ceph cli invocation is able to use it to connect to the cluster.
Signed-off-by: Sage Weil <sage@redhat.com>
* refs/pull/27490/head:
PendingReleaseNotes: note on python3.6 changes
Reviewed-by: Boris Ranto <branto@redhat.com>
Reviewed-by: Sage Weil <sage@redhat.com>
so user can install python3-rados, instead of python36-rados, without
specifying the minor version of python. also, we should not break our
teuthology tests with this naming scheme change. for instance, our
cephfs qa suite installs `python3-cephfs` for testing the `cephfs-shell`
Signed-off-by: Kefu Chai <kchai@redhat.com>
rgw: Evaluating bucket policies also while reading permissions for an…
Reviewed-by: Adam C. Emerson <aemerson@redhat.com>
Reviewed-by: Matt Benjamin <mbenjamin@redhat.com>
* refs/pull/27051/head:
mds: open import bounding dirfrags in batch
mds: remove superfluous error in StrayManager::advance_delayed()
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
The dashboards Prometheus receiver API needed to receive a POST from the
frontend, but POSTs aren't allowed by default by any read only
users. As a result the receiver API call had thrown a 403 error which
redirected the user to the 403 error page. Now you can get the last
notifications via GET. This prevents the redirection for read only
users, as a result they can get the last notifications and also
see all other allowed pages again.
Fixes: https://tracker.ceph.com/issues/39086
Signed-off-by: Stephan Müller <smueller@suse.com>
After PR https://github.com/ceph/ceph/pull/26572, when RGW is not
configured, accessing /rgw drop-down (daemons, users or buckets)
results in nothing apparently happening (not even an error).
Under the curtains, what is happening is that the ModuleStatusGuard
has redirected the route to the rgw/501, but as this route is now
under parent rgw route handler, which sets CanActivateChild guards,
this results in a new ModuleStatusGuard invokation, a subsequent
failure and a new redirection to rgw/501.
Several approaches could be taken here:
- Remove error pages from lazy-loaded modules. Probably it does not
make sense to have a 501 page per component.
- Add some whitelist to avoid this kind of loop (e.g.: 501, or any
error page).
- Set a max number of redirections (cautionary measure).
Fixes: https://tracker.ceph.com/issues/39125
Signed-off-by: Ernesto Puerta <epuertat@redhat.com>
* common/auth_handler.h: add an abstract class of AuthHandler, the one who is interested in
an authenticated peer should implement this class
* mon/MonClient: let mon::Client implement AuthServer, as it has access the keyring. it
will update the registered AuthHandler if the client (peer) is
authenticated.
* osd: implement AuthHandler class. we will keep track of the connected
sessions along their caps in a follow-up change.
Signed-off-by: Kefu Chai <kchai@redhat.com>
crimson/osd: consolidate the code to initialize msgrs
Reviewed-by: Samuel Just <sjust@redhat.com>
Reviewed-by: Radoslaw Zarzynski <rzarzyns@redhat.com>
* refs/pull/27256/head:
mds: drop reconnect message from non-existent session
Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
Reviewed-by: Zheng Yan <zyan@redhat.com>
