This patch is based on changes proposed by Mark Barnes
<mark.barnes@ocado.com> (lperiquito) in commit
af71f6a518.
Signed-off-by: Mark Barnes <mark.barnes@ocado.com>
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
This patch is based on changes proposed by Mark Barnes
<mark.barnes@ocado.com> (lperiquito) in commit
af71f6a518.
Signed-off-by: Mark Barnes <mark.barnes@ocado.com>
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
The change has been extracted from changeset proposed
in commit af71f6a518 by
Mark Barnes <mark.barnes@ocado.com> (lperiquito).
Signed-off-by: Mark Barnes <mark.barnes@ocado.com>
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
The change has been extracted from changeset proposed
in commit af71f6a518 by
Mark Barnes <mark.barnes@ocado.com> (lperiquito).
Signed-off-by: Mark Barnes <mark.barnes@ocado.com>
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
The change has been extracted from changeset proposed
in commit af71f6a518 by
Mark Barnes <mark.barnes@ocado.com> (lperiquito).
Signed-off-by: Mark Barnes <mark.barnes@ocado.com>
Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
the symbols of buffer::list::iterator_impl<> were wrongly exposed
in previous infernalis release, and the clients linked against
librados are very likely using them. so we need to document this
change.
Signed-off-by: Kefu Chai <kchai@redhat.com>
buffer::list::iterator_impl symbols are referenced by const_iterator
and iterator, and are exposed as weak symbols. if a source file is
compiled using the buffer.h, the produced object file will reference
these symbols as well, so we'd better hiding them and avoid using
them in the header file.
as a side-effect, buffer::list::const_iterator is also hidden, but
currently we don't have any librados client using this class, so
we can just leave it as an internal class at this moment.
Fixes: #14788
Signed-off-by: Kefu Chai <kchai@redhat.com>
When journaling is enabled, a read op should acquire the exclusive
lock and replay the journal in case uncommitted writes are stored
in the journal.
Signed-off-by: Jason Dillaman <dillaman@redhat.com>
rgw multisite v2 (Yehuda Sadeh, Orit Wasserman, Casey Bodley)
- rename region to zonegroup
- new zone and zonegroup configuration
- new realm and period structures
- rgw handles synchronization (no need for sync agent)
- active-active data
Signed-off-by: Yehuda Sadeh
When we send in an unrecognized field for set-quota, specify that the
options should be `max_bytes` or `max_objects`. This doesn't affect the
ceph-cli as the choices are caught at an earlier stage, however the
error message is seen while reaching the cluster via rados api for eg.
Signed-off-by: Abhishek Lekshmanan <abhishek@suse.com>
Don't try to sync an object with empty name (shouldn't happen, but
bugs may happen). Also, don't try to write_meta() of an empty object
name, in case we get there through a different path somehow.
Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
RGWMetaSyncCR uses a period history Cursor to track its position. it
uses this to get the max sync markers for each shard from the following
period, so that RGWMetaSyncShardCR knows when to stop syncing and return
control to RGWMetaSyncCR
Signed-off-by: Casey Bodley <cbodley@redhat.com>
