See bug #2462.
The following sequence could cause a log assuming a non-empty pg
to an empty replica:
1. primary sends query to stray
2. stray sends notify to primary
3. primary sends purge to stray removing stray from peer_info
4. stray recieves query and sends a notify
5. stray recieves purge and purges its pg
6. primary recieves notify from stray and adds it to peer_info
note: peer_info[stray] is now wrong
7. acting set changes, primary is still primary, stray is replica
8. primary sends log to replica based on incorrect info from 6.
This patch adds a purged_peer set which is populated during purge_strays
and cleared during start_peering_interval. The primary will ignore
notifies from the peer once the peer is in this set.
Signed-off-by: Samuel Just <sam.just@inktank.com>
Check the authenticated peer name directly; don't store a copy in the
OSDCaps struct. Also set the entity_name member of the session!
Signed-off-by: Sage Weil <sage@inktank.com>
We depend on leveldb, which only builds on
amd64 armel armhf i386 ia64 mipsel
Reported-by: Laszlo Boszormenyi (GCS) <gcs@debian.hu>
Signed-off-by: Sage Weil <sage@inktank.com>
This introduces two new config variables, osd_crush_location
and osd_crush_weight. Not currently included in config_opts.h,
as these are not used in the C++ code.
Signed-off-by: Tommi Virtanen <tv@inktank.com>
If we encounter a 'key=' with no value, clear any previous value for that
key, so that
'a=foo b=bar a='
is equivalent to
'b=bar'
Fixes: #2540
Signed-off-by: Sage Weil <sage@inktank.com>
* "user gen" is gone and is now "user create"
* --uid option is mandatory for "user create"
Mailing list thread: http://www.spinics.net/lists/ceph-devel/msg06744.html
Signed-off-by: Florian Haas <florian@hastexo.com>
- radosgw-admin usage show
- radosgw-admin usage trim
Both commands can get --start-date, --end-date, --uid
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
New config options for usage logging:
- rgw_enable_usage_log: enable usage logging
- rgw_usage_log_flush_threshold - limit on number of pending updates
before synchronously flushing update
- rgw_usage_log_tick_interval - asynchronous flush interval
- rgw_usage_max_shards - split info across that many objects
- rgw_usage_max_user_shards - split single user info across that many
objects
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
The new methods are:
- user_usage_log_add: add new usage information
- user_usage_log_read: get usage information
- user_usage_log_trim: remove usage information
Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
This is useful for reporting which features an osd supports, and for
testing rados_exec. Update the rados api tests to use this method
instead of test_exec, which was removed.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
Reviewed-by: Sage Weil <sage@inktank.com>
PGMap->num_pg_by_state is a PG state to number of PG in the state
mapping. PGMonitor::update_logger wrongly interprets the mapping.
Signed-off-by: Yan, Zheng <zheng.z.yan@intel.com>
Signed-off-by: Sage Weil <sage@inktank.com>
The actual data object ids don't need to be artificially restricted in
length. RBD_MAX_BLOCK_NAME_SIZE just limits the size of the object
prefix, since it's used in rbd_info_t.
Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
