ceph (0.22.1-1) experimental; urgency=low

  * osd: fix hang during mkfs journal creation
  * objecter: fix rare hang during shutdown
  * msgr: fix reconnect/disconnect errors due to timeouts
  * init-ceph: check for correct instance in daemon_is_running()
  * filestore: deliberate crash on ENOSPC or EIO to avoid corruption
  * filestore: split xattrs into chunk (partial workaround for xattr size)
  * filestore: ignore ENOSPC on xattrs until workaround for extN is ready
  * radosgw: return 204 on removal of bucket/object

 -- Sage Weil <sage@newdream.net>  Fri, 22 Oct 2010 16:18:49 -0700

ceph (0.22-4) experimental; urgency=low

  * more packaging fixes 

 -- Sage Weil <sage@newdream.net>  Tue, 19 Oct 2010 09:00:21 -0700

ceph (0.22-3) experimental; urgency=low

  * packaging fixes 

 -- Sage Weil <sage@newdream.net>  Mon, 18 Oct 2010 10:23:02 -0700

ceph (0.22-1) experimental; urgency=low

  * mon: improved monitor naming
  * osd: support seprate interface for cluster-internal traffic
  * mds: many many snapshot, cap protocol fixes
  * mon: back off on clock drift errors/warnings
  * mds: fix cow on xattrs
  * mds: many recurstat stat fixes
  * uclient/cfuse: improved caching and readdir fixes
  * mon: tri laggy standby MDS's from map
  * osd: data integrity fixes
  * trim stale connections
  * osd: fix hang during mkfs
  * improved debug log on SIGSEG
  * uclient/cfuse: cache readdir results
  * mon: improved osd failure smarts
  * mds/osd: tcmalloc (less memory utilization)
  * mds: set file layout/striping on per-directory/subtree basis
  * cfuse: mount server subdir
  * cfuse: log to file if run as daemon
  * allow monitor hostnames on command line
  * mon: ability to blacklist/unblacklist clients for osd cluster
  * cdebugpack: add debug capture tool
  * mds: respawn on suicide
  * msgr: more efficient reconnect resync
  * standard date/time format

 -- Sage Weil <sage@newdream.net>  Fri, 15 Oct 2010 13:42:24 -0700

ceph (0.21.3-1) unstable; urgency=low

  * cfuse: fixed readdir for large directories
  * cfuse: fixed truncation (now passes fsx)
  * radosgw: fixed information leakage between sessions
  * radosgw: support wildcard dns for bucketname.domain/object access
  * mds: fixed snap bug with directory renames
  * mon: less log noise
  * osdmap: allowed blacklisting of entire ips
  * mds: fix journal replay bug with reconnecting clients
  * mds: fix journal replay bug with inconsistent lock state, subsequent crash 

 -- Sage Weil <sage@newdream.net>  Fri, 17 Sep 2010 15:44:06 -0700

ceph (0.21.2-1) unstable; urgency=low

  * osd: less log noise
  * osd: mark down old heartbeat peers
  * filejournal: clean up init sequence, fewer confusing errors on startup
  * msgr: fix throttler leak when closing connections (fixes deadlock)
  * osdmaptool: don't crash on corrupt input
  * crush: error out instead of crashing on ambiguous hierarchy
  * mds: error to client on invalid op
  * mds: fix ENOTEMPTY checks on rmdir
  * osd: fix race between reads and cloned objects
  * auth: fix keyring load search when $HOME not defined
  * client: fix xattr cap writeback
  * client: fix snap vs metadata writeback
  * osd: fix journal, btrfs throttling
  * msgr: fix memory leak on closed connections

 -- Sage Weil <sage@newdream.net>  Thu, 26 Aug 2010 15:22:42 -0700

ceph (0.21.1-1) unstable; urgency=high

  * ceph.spec: many fixes
  * mds mds->client message crash
  * monmaptool: man page typo
  * rados: usage fix
  * osd: fix heartbeat to/from node
  * osd: fix replies to dup/already committed requests
  * librados: hpp fix
  * mds: fix cap/snaprealm behavior on readdir
  * debian: fix fuse-utils depcs
  * cclass: fix .so loading
  * osd: fix crazy slowdown during recovery/repeering due to hadnle_osd_map
  * cauthtool: fix man page example for fs clients
  * mkcephfs to /sbin
  * fix log rotation

 -- Sage Weil <sage@fatty.ops.newdream.net>  Wed, 11 Aug 2010 14:55:25 -0700

ceph (0.21-1) unstable; urgency=low

  * improved logging infrastructure
  * log rotation
  * mkfs improvements
  * rbd tool, class (rados block device)
  * mds: return ENOTEMPTY when removing directory with snapshots
  * mds: lazy io support
  * msgr: send messages directly to connection handle (w/o looking up
    destination address)
  * faster atomic via libatomic-ops
  * cfuse now stable
  * mon: recovery improvements (e.g., when one mon is down for a long time)
  * mon: warn on monitor clock drift
  * osd: large object support 
  * osd: heartbeat fixes
  * osd: journaling fixes, improvements (bugs, better use of direct io)
  * osd: snapshot rollback op (for rbd)
  * radosgw fixes, improvements
  * many memory leak fixes

 -- Sage Weil <sage@newdream.net>  Tue, 27 Jul 2010 12:35:09 -0700

ceph (0.20.2-1) unstable; urgency=low

  * Initial release (Closes: #506040)

 -- sage <sage@newdream.net>  Tue, 20 Jul 2010 19:40:44 -0800