v10.2.3 Jewel Released

Sep 28, 2016 TheAnalyst

This point release fixes several important bugs in RBD mirroring, RGW multi-site, CephFS, and RADOS.

We recommend that all v10.2.x users upgrade.

For more detailed information, see the complete changelog.

Notable Changes ¶

build/ops: 60-ceph-partuuid-workaround-rules still needed by debian jessie (udev 215-17) (issue#16351, pr#10653, runsisi, Loic Dachary)
build/ops: ceph Resource Agent does not work with systemd (issue#14828, pr#9917, Nathan Cutler)
build/ops: ceph-base requires parted (issue#16095, pr#10008, Ken Dreyer)
build/ops: ceph-osd-prestart.sh contains Upstart-specific code (issue#15984, pr#10364, Nathan Cutler)
build/ops: mount.ceph: move from ceph-base to ceph-common and add symlink in /sbin for SUSE (issue#16598, issue#16645, pr#10357, Nathan Cutler, Dan Horák, Ricardo Dias, Kefu Chai)
build/ops: need rocksdb commit 7ca731b12ce for ppc64le build (issue#17092, pr#10816, Nathan Cutler)
build/ops: rpm: OBS needs ExclusiveArch (issue#16936, pr#10614, Michel Normand)
cli: ceph command line tool chokes on ceph –w (the dash is unicode ‘en dash’ &ndash, copy-paste to reproduce) (issue#12287, pr#10420, Oleh Prypin, Kefu Chai)
common: expose buffer const_iterator symbols (issue#16899, pr#10552, Noah Watkins)
common: global-init: fixup chown of the run directory along with log and asok files (issue#15607, pr#8754, Karol Mroz)
fs: ceph-fuse: link to libtcmalloc or jemalloc (issue#16655, pr#10303, Yan, Zheng)
fs: client: crash in unmount when fuse_use_invalidate_cb is enabled (issue#16137, pr#10106, Yan, Zheng)
fs: client: fstat cap release (issue#15723, pr#9562, Yan, Zheng, Noah Watkins)
fs: essential backports for OpenStack Manila (issue#15406, issue#15614, issue#15615, pr#10453, John Spray, Ramana Raja, Xiaoxi Chen)
fs: fix double-unlock on shutdown (issue#17126, pr#10847, Greg Farnum)
fs: fix mdsmap print_summary with standby replays (issue#15705, pr#9547, John Spray)
fs: fuse mounted file systems fails SAMBA CTDB ping_pong rw test with v9.0.2 (issue#12653, issue#15634, pr#10108, Yan, Zheng)
librados: Missing export for rados_aio_get_version in src/include/rados/librados.h (issue#15535, pr#9574, Jim Wright)
librados: osd: bad flags can crash the osd (issue#16012, pr#9997, Sage Weil)
librbd: Close journal and object map before flagging exclusive lock as released (issue#16450, pr#10053, Jason Dillaman)
librbd: Crash when utilizing advisory locking API functions (issue#16364, pr#10051, Jason Dillaman)
librbd: ExclusiveLock object leaked when switching to snapshot (issue#16446, pr#10054, Jason Dillaman)
librbd: FAILED assert(object_no < m_object_map.size()) (issue#16561, pr#10647, Jason Dillaman)
librbd: Image removal doesn’t necessarily clean up all rbd_mirroring entries (issue#16471, pr#10009, Jason Dillaman)
librbd: Object map/fast-diff invalidated if journal replays the same snap remove event (issue#16350, pr#10010, Jason Dillaman)
librbd: Timeout sending mirroring notification shouldn’t result in failure (issue#16470, pr#10052, Jason Dillaman)
librbd: Whitelist EBUSY error from snap unprotect for journal replay (issue#16445, pr#10055, Jason Dillaman)
librbd: cancel all tasks should wait until finisher is done (issue#16517, pr#9752, Haomai Wang)
librbd: delay acquiring lock if image watch has failed (issue#16923, pr#10827, Jason Dillaman)
librbd: fix missing return statement if failed to get mirror image state (issue#16600, pr#10144, runsisi)
librbd: flag image as updated after proxying maintenance op (issue#16404, pr#9883, Jason Dillaman)
librbd: mkfs.xfs slow performance with discards and object map (issue#16707, issue#16689, pr#10649, Jason Dillaman)
librbd: potential use after free on refresh error (issue#16519, pr#9952, Mykola Golub)
librbd: rbd-nbd does not properly handle resize notifications (issue#15715, pr#10679, Mykola Golub)
librbd: the option ‘rbd_cache_writethrough_until_flush=true’ dosn’t work (issue#16740, issue#16386, issue#16708, issue#16654, issue#16478, pr#10797, Mykola Golub, xinxin shu, Xiaowei Chen, Jason Dillaman)
mds: tell command blocks forever with async messenger (TestVolumeClient.test_evict_client failure) (issue#16288, pr#10501, Douglas Fuller)
mds: Confusing MDS log message when shut down with stalled journaler reads (issue#15689, pr#9557, John Spray)
mds: Deadlock on shutdown active rank while busy with metadata IO (issue#16042, pr#10502, Patrick Donnelly)
mds: Failing file operations on kernel based cephfs mount point leaves unaccessible file behind on hammer 0.94.7 (issue#16013, pr#10199, Yan, Zheng)
mds: Fix shutting down mds timed-out due to deadlock (issue#16396, pr#10500, Zhi Zhang)
mds: MDSMonitor fixes (issue#16136, pr#9561, xie xingguo)
mds: MDSMonitor::check_subs() is very buggy (issue#16022, pr#10103, Yan, Zheng)
mds: Session::check_access() is buggy (issue#16358, pr#10105, Yan, Zheng)
mds: StrayManager.cc: 520: FAILED assert(dnl->is_primary()) (issue#15920, pr#9559, Yan, Zheng)
mds: enforce a dirfrag limit on entries (issue#16164, pr#10104, Patrick Donnelly)
mds: fix SnapRealm::have_past_parents_open() (issue#16299, pr#10499, Yan, Zheng)
mds: fix getattr starve setattr (issue#16154, pr#9560, Yan, Zheng)
mds: wrongly treat symlink inode as normal file/dir when symlink inode is stale on kcephfs (issue#15702, pr#9405, Zhi Zhang)
mon: “mon metadata” fails when only one monitor exists (issue#15866, pr#10654, John Spray, Kefu Chai)
mon: Monitor: validate prefix on handle_command() (issue#16297, pr#10036, You Ji)
mon: OSDMonitor: drop pg temps from not the current primary (issue#16127, pr#9998, Samuel Just)
mon: prepare_pgtemp needs to only update up_thru if newer than the existing one (issue#16185, pr#10001, Samuel Just)
msgr: AsyncConnection::lockmsg/async lockdep cycle: AsyncMessenger::lock, MDSDaemon::mds_lock, AsyncConnection::lock (issue#16237, pr#10004, Haomai Wang)
msgr: async messenger mon crash (issue#16378, issue#16418, pr#9996, Haomai Wang)
msgr: backports of all asyncmsgr fixes to jewel (issue#15503, issue#15372, pr#9633, Yan Jun, Haomai Wang, Piotr Dałek)
msgr: msg/async: connection race hang (issue#15849, pr#10003, Haomai Wang)
osd: FileStore: umount hang because sync thread doesn’t exit (issue#15695, pr#9105, Kefu Chai)
osd: Fixes for list-inconsistent-* (issue#15766, issue#16192, issue#15719, pr#9565, David Zafman)
osd: New pools have bogus stuck inactive/unclean HEALTH_ERR messages until they are first active and clean (issue#14952, pr#10007, Sage Weil)
osd: OSD crash with Hammer to Jewel Upgrade: void FileStore::init_temp_collections() (issue#16672, pr#10561, David Zafman)
osd: OSD failed to subscribe skipped osdmaps after ceph osd pause (issue#17023, pr#10804, Kefu Chai)
osd: ObjectCacher split BufferHead read fix (issue#16002, pr#10074, Greg Farnum)
osd: ReplicatedBackend doesn’t increment stats on pull, only push (issue#16277, pr#10421, Kefu Chai)
osd: Scrub error: 0/1 pinned (issue#15952, pr#9576, Samuel Just)
osd: crash adding snap to purged_snaps in ReplicatedPG::WaitingOnReplicas (issue#15943, pr#9575, Samuel Just)
osd: partprobe intermittent issues during ceph-disk prepare (issue#15176, pr#10497, Marius Vollmer, Loic Dachary)
osd: saw valgrind issues in ReplicatedPG::new_repop (issue#16801, pr#10760, Kefu Chai)
osd: sparse_read on ec pool should return extends with correct offset (issue#16138, pr#10006, kofiliu)
osd:sched_time not actually randomized (issue#15890, pr#9578, xie xingguo)
rbd: ImageReplayer::is_replaying does not include flush state (issue#16970, pr#10790, Jason Dillaman)
rbd: Journal duplicate op detection can cause lockdep error (issue#16363, pr#10044, Jason Dillaman)
rbd: Journal needs to handle duplicate maintenance op tids (issue#16362, pr#10045, Jason Dillaman)
rbd: Unable to disable journaling feature if in unexpected mirror state (issue#16348, pr#10042, Jason Dillaman)
rbd: bashism in src/rbdmap (issue#16608, pr#10786, Jason Dillaman)
rbd: doc: format 2 now is the default image format (issue#17026, pr#10732, Chengwei Yang)
rbd: hen journaling is enabled, a flush request shouldn’t flush the cache (issue#15761, pr#10041, Yuan Zhou)
rbd: possible race condition during journal transition from replay to ready (issue#16198, pr#10047, Jason Dillaman)
rbd: qa/workunits/rbd: respect RBD_CREATE_ARGS environment variable (issue#16289, pr#9721, Mykola Golub)
rbd: rbd-mirror should disable proxied maintenance ops for non-primary image (issue#16411, pr#10050, Jason Dillaman)
rbd: rbd-mirror: FAILED assert(m_local_image_ctx->object_map != nullptr) (issue#16558, pr#10646, Jason Dillaman)
rbd: rbd-mirror: FAILED assert(m_on_update_status_finish == nullptr) (issue#16956, pr#10792, Jason Dillaman)
rbd: rbd-mirror: FAILED assert(m_state == STATE_STOPPING) (issue#16980, pr#10791, Jason Dillaman)
rbd: rbd-mirror: ensure replay status formatter has completed before stopping replay (issue#16352, pr#10043, Jason Dillaman)
rbd: rbd-mirror: include local pool id in resync throttle unique key (issue#16536, issue#15239, issue#16488, issue#16491, issue#16329, issue#15108, issue#15670, pr#10678, Ricardo Dias, Jason Dillaman)
rbd: rbd-mirror: potential race condition accessing local image journal (issue#16230, pr#10046, Jason Dillaman)
rbd: rbd-mirror: reduce memory footprint during journal replay (issue#16321, issue#16489, issue#16622, issue#16539, issue#16223, issue#16349, pr#10684, Mykola Golub, Jason Dillaman)
rgw: A query on a static large object fails with 404 error (issue#16015, pr#9544, Radoslaw Zarzynski)
rgw: Add zone rename to radosgw_admin (issue#16934, pr#10663, Shilpa Jagannath)
rgw: Bucket index shards orphaned after bucket delete (issue#16412, pr#10525, Orit Wasserman)
rgw: Bug when using port 443s in rgw. (issue#16548, pr#10664, Pritha Srivastava)
rgw: Fallback to Host header for bucket name. (issue#15975, pr#10693, Robin H. Johnson)
rgw: Fix civetweb IPv6 (issue#16928, pr#10580, Robin H. Johnson)
rgw: Increase log level for messages occuring while running rgw admin command (issue#16935, pr#10765, Shilpa Jagannath)
rgw: No Last-Modified, Content-Size and X-Object-Manifest headers if no segments in DLO manifest (issue#15812, pr#9265, Radoslaw Zarzynski)
rgw: RGWPeriodPuller tries to pull from itself (issue#16939, pr#10764, Casey Bodley)
rgw: Set Access-Control-Allow-Origin to a Asterisk if allowed in a rule (issue#15348, pr#9453, Wido den Hollander)
rgw: Swift API returns double space usage and objects of account metadata (issue#16188, pr#10148, Albert Tu)
rgw: account/container metadata not actually present in a request are deleted during POST through Swift API (issue#15977, issue#15779, pr#9542, Radoslaw Zarzynski)
rgw: add socket backlog setting for via ceph.conf (issue#16406, pr#10216, Feng Guo)
rgw: add tenant support to multisite sync (issue#16469, issue#16121, issue#16665, pr#10845, Yehuda Sadeh, Josh Durgin, Casey Bodley, Pritha Srivastava)
rgw: add_zone only clears master_zone if –master=false (issue#15901, pr#9327, Casey Bodley)
rgw: aws4 parsing issue (issue#15940, issue#15939, pr#9545, Yehuda Sadeh)
rgw: aws4: add STREAMING-AWS4-HMAC-SHA256-PAYLOAD support (issue#16146, pr#10167, Radoslaw Zarzynski, Javier M. Mellid)
rgw: backport merge of static sites fixes (issue#15555, issue#15532, issue#15531, pr#9568, Robin H. Johnson)
rgw: can set negative max_buckets on RGWUserInfo (issue#14534, pr#10655, Yehuda Sadeh)
rgw: cleanup radosgw-admin temp command as it was deprecated (issue#16023, pr#9390, Vikhyat Umrao)
rgw: comparing return code to ERR_NOT_MODIFIED in rgw_rest_s3.cc (needs minus sign) (issue#16327, pr#9790, Nathan Cutler)
rgw: custom metadata aren’t camelcased in Swift’s responses (issue#15902, pr#9267, Radoslaw Zarzynski)
rgw: data sync stops after getting error in all data log sync shards (issue#16530, pr#10073, Yehuda Sadeh)
rgw: default zone and zonegroup cannot be added to a realm (issue#16839, pr#10658, Casey Bodley)
rgw: document multi tenancy (issue#16635, pr#10217, Pete Zaitcev)
rgw: don’t unregister request if request is not connected to manager (issue#15911, pr#9242, Yehuda Sadeh)
rgw: failed to create bucket after upgrade from hammer to jewel (issue#16627, pr#10524, Orit Wasserman)
rgw: fix ldap bindpw parsing (issue#16286, pr#10518, Matt Benjamin)
rgw: fix multi-delete query param parsing. (issue#16618, pr#10188, Robin H. Johnson)
rgw: improve support for Swift’s object versioning. (issue#15925, pr#10710, Radoslaw Zarzynski)
rgw: initial slashes are not properly handled in Swift’s BulkDelete (issue#15948, pr#9316, Radoslaw Zarzynski)
rgw: master: build failures with boost > 1.58 (issue#16392, issue#16391, pr#10026, Abhishek Lekshmanan)
rgw: multisite segfault on ~RGWRealmWatcher if realm was deleted (issue#16817, pr#10660, Casey Bodley)
rgw: multisite sync races with deletes (issue#16222, issue#16464, issue#16220, issue#16143, pr#10293, Yehuda Sadeh, Casey Bodley)
rgw: multisite: preserve zone’s extra pool (issue#16712, pr#10537, Abhishek Lekshmanan)
rgw: object expirer’s hints might be trimmed without processing in some circumstances (issue#16705, issue#16684, pr#10763, Radoslaw Zarzynski)
rgw: radosgw-admin failure for user create after upgrade from hammer to jewel (issue#15937, pr#9294, Orit Wasserman, Abhishek Lekshmanan)
rgw: radosgw-admin: EEXIST messages for create operations (issue#15720, pr#9268, Abhishek Lekshmanan)
rgw: radosgw-admin: inconsistency in uid/email handling (issue#13598, pr#10520, Matt Benjamin)
rgw: realm pull fails when using apache frontend (issue#15846, pr#9266, Orit Wasserman)
rgw: retry on bucket sync errors (issue#16108, pr#9425, Yehuda Sadeh)
rgw: s3website: x-amz-website-redirect-location header returns malformed HTTP response (issue#15531, pr#9099, Robin H. Johnson)
rgw: segfault in RGWOp_MDLog_Notify (issue#16666, pr#10662, Casey Bodley)
rgw: segmentation fault on error_repo in data sync (issue#16603, pr#10523, Casey Bodley)
rgw: selinux denials in RGW (issue#16126, pr#10519, Boris Ranto)
rgw: support size suffixes for –max-size in radosgw-admin command (issue#16004, pr#9743, Vikhyat Umrao)
rgw: updating CORS/ACLs might not work in some circumstances (issue#15976, pr#9543, Radoslaw Zarzynski)
rgw: use zone endpoints instead of zonegroup endpoints (issue#16834, pr#10659, Casey Bodley)
tests: improve rbd-mirror test case coverage (issue#16197, pr#9631, Mykola Golub, Jason Dillaman)
tests: rados/test.sh workunit timesout on OpenStack (issue#15403, pr#8904, Loic Dachary)
tools: ceph-disk: Accept bcache devices as data disks (issue#13278, pr#8497, Peter Sabaini)
tools: rados: Add cleanup message with time to rados bench output (issue#15704, pr#9740, Vikhyat Umrao)
tools: src/script/subman fails with KeyError: ‘nband’ (issue#16961, pr#10625, Loic Dachary, Ali Maredia)