doc: rgw_region_root_pool option should be in [global]
Reviewed-by: Abhishek Lekshmanan <abhishek@suse.com>
Reviewed-by: Yehuda Sadeh <yehuda@redhat.com>
The python scripts are not yet compatible with python3, yet it is the
default on jessie. Force the creation of the virtualenv to use python2.7
instead. The wheelhouse is already explicitly populated for both python3
and python2.7 by install-deps.sh, regardless of the default interpreter.
Signed-off-by: Loic Dachary <loic@dachary.org>
We need to update the high marker even if we skip entries, and eventually
flush it. This is needed so that our position in the bucket index log that
we follow is reflected correctly.
Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
Need to update the data log anyway, so that it reflects that there were
changes in the bucket index log.
Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
Do not use C struct initilization style to avoid eclipse errors.
We also don't use c++ struct initalize as members are not specified
and it just happens that in latest accelio version one of the fields
has changed in the middle of the struct.
Signed-off-by: Roi Dayan <roid@mellanox.com>
They will split from the parent on their own, and the OSD drops these
PGs on the floor anyway in OSD::handle_pg_create().
Signed-off-by: Sage Weil <sage@redhat.com>
Fixes#15243
When removing a plain null versioned object (was created prior to bucket versioning
enabled), need to convert the bucket index representation to a versioned one. This
is needed so that all the versioning mechanics play together.
Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
make CEPH_BUILD_VIRTUALENV=/tmp TESTS=ceph-detect-init/run-tox.sh check
sets the CEPH_DETECT_INIT_VIRTUALENV environment variable
CEPH_BUILD_VIRTUALENV=/tmp make TESTS=ceph-detect-init/run-tox.sh check
does not because Makefile-env.am overrides it.
Do not set CEPH_BUILD_VIRTUALENV in Makefile-env.am so the variable from
the environment can be used. Require that CEPH_BUILD_VIRTUALENV has a
trailing slash so that the default is just an empty variable.
Signed-off-by: Loic Dachary <loic@dachary.org>
We remember the range of requested full map in requested_full_first/last
and prevent sending duplicate requests.
But monitor will cap the reply to osd_map_message_max number of maps, for example,
OSD request [100, 200] while monitor only return [100,149], previous code think
[150, 200] is dup and prevent the OSD to send out the request, which is wrong.
Fix this by clear the requested_full_first/last field at the end of handle_osd_map.
Fixes: #15130
Signed-off-by: Xiaoxi Chen <xiaoxchen@ebay.com>
Theoretically even if _have_pg() returns ture, we still can't assert that
_lookup_lock_pg() will always succeed. This is because when we switch between
these two methods, we will drop pg_map_lock, and thus may let a pg removal
sneak in, which may eventually cause divergence.
However this is a really rare case, and is less likely to happen in a
production environment. But this pr provided a safer way to achieve
the same goal and is a little faster by eliminating a duplicated search
from the pg_map, which makes it meaningful.
Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
On returning false, the require_mon_peer() will internally decrease
the reference of the input message.
So the put() method here is duplicated and will cause reference
underflow and thus need to be dropped.
Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
