From mboxrd@z Thu Jan 1 00:00:00 1970 From: Dan van der Ster Subject: Re: v0.94.6 Hammer released Date: Wed, 24 Feb 2016 10:31:01 +0100 Message-ID: References: Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from mail-lb0-f175.google.com ([209.85.217.175]:34656 "EHLO mail-lb0-f175.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752529AbcBXJbo convert rfc822-to-8bit (ORCPT ); Wed, 24 Feb 2016 04:31:44 -0500 Received: by mail-lb0-f175.google.com with SMTP id of3so6903119lbc.1 for ; Wed, 24 Feb 2016 01:31:43 -0800 (PST) Received: from mail-lb0-f174.google.com (mail-lb0-f174.google.com. [209.85.217.174]) by smtp.gmail.com with ESMTPSA id r191sm254329lfe.28.2016.02.24.01.31.41 for (version=TLSv1/SSLv3 cipher=OTHER); Wed, 24 Feb 2016 01:31:41 -0800 (PST) Received: by mail-lb0-f174.google.com with SMTP id of3so6902608lbc.1 for ; Wed, 24 Feb 2016 01:31:41 -0800 (PST) In-Reply-To: Sender: ceph-devel-owner@vger.kernel.org List-ID: To: Sage Weil Cc: "ceph-devel@vger.kernel.org" , ceph-users Thanks Sage, looking forward to some scrub randomization. Were binaries built for el6? http://download.ceph.com/rpm-hammer/el6/x8= 6_64/ Cheers, Dan On Tue, Feb 23, 2016 at 5:01 PM, Sage Weil wrote: > This Hammer point release fixes a range of bugs, most notably a fix f= or > unbounded growth of the monitor=E2=80=99s leveldb store, and a workar= ound in the > OSD to keep most xattrs small enough to be stored inline in XFS inode= s. > > We recommend that all hammer v0.94.x users upgrade. > > For more detailed information, see the complete changelog: > > http://docs.ceph.com/docs/master/_downloads/v0.94.6.txt > > Notable Changes > --------------- > > * build/ops: Ceph daemon failed to start, because the service name wa= s already used. (#13474, Chuanhong Wang) > * build/ops: LTTng-UST tracing should be dynamically enabled (#13274,= Jason Dillaman) > * build/ops: ceph upstart script rbdmap.conf incorrectly processes pa= rameters (#13214, Sage Weil) > * build/ops: ceph.spec.in License line does not reflect COPYING (#129= 35, Nathan Cutler) > * build/ops: ceph.spec.in libcephfs_jni1 has no %post and %postun (#= 12927, Owen Synge) > * build/ops: configure.ac: no use to add "+" before ac_ext=3Dc (#1433= 0, Kefu Chai, Robin H. Johnson) > * build/ops: deb: strip tracepoint libraries from Wheezy/Precise buil= ds (#14801, Jason Dillaman) > * build/ops: init script reload doesn't work on EL7 (#13709, Herv=C3=A9= Rousseau) > * build/ops: init-rbdmap uses distro-specific functions (#12415, Bori= s Ranto) > * build/ops: logrotate reload error on Ubuntu 14.04 (#11330, Sage Wei= l) > * build/ops: miscellaneous spec file fixes (#12931, #12994, #12924, #= 12360, Boris Ranto, Nathan Cutler, Owen Synge, Travis Rhoden, Ken Dreye= r) > * build/ops: pass tcmalloc env through to ceph-os (#14802, Sage Weil) > * build/ops: rbd-replay-* moved from ceph-test-dbg to ceph-common-dbg= as well (#13785, Loic Dachary) > * build/ops: unknown argument --quiet in udevadm settle (#13560, Jaso= n Dillaman) > * common: Objecter: pool op callback may hang forever. (#13642, xie x= ingguo) > * common: Objecter: potential null pointer access when do pool_snap_l= ist. (#13639, xie xingguo) > * common: ThreadPool add/remove work queue methods not thread safe (#= 12662, Jason Dillaman) > * common: auth/cephx: large amounts of log are produced by osd (#1361= 0, Qiankun Zheng) > * common: client nonce collision due to unshared pid namespaces (#130= 32, Josh Durgin) > * common: common/Thread:pthread_attr_destroy(thread_attr) when done w= ith it (#12570, Piotr Da=C5=82ek) > * common: log: Log.cc: Assign LOG_DEBUG priority to syslog calls (#13= 993, Brad Hubbard) > * common: objecter: cancellation bugs (#13071, Jianpeng Ma) > * common: pure virtual method called (#13636, Jason Dillaman) > * common: small probability sigabrt when setting rados_osd_op_timeout= (#13208, Ruifeng Yang) > * common: wrong conditional for boolean function KeyServer::get_auth(= ) (#9756, #13424, Nathan Cutler) > * crush: crash if we see CRUSH_ITEM_NONE in early rule step (#13477, = Sage Weil) > * doc: man: document listwatchers cmd in "rados" manpage (#14556, Kef= u Chai) > * doc: regenerate man pages, add orphans commands to radosgw-admin(8)= (#14637, Ken Dreyer) > * fs: CephFS restriction on removing cache tiers is overly strict (#1= 1504, John Spray) > * fs: fsstress.sh fails (#12710, Yan, Zheng) > * librados: LibRadosWatchNotify.WatchNotify2Timeout (#13114, Sage Wei= l) > * librbd: ImageWatcher shouldn't block the notification thread (#1437= 3, Jason Dillaman) > * librbd: diff_iterate needs to handle holes in parent images (#12885= , Jason Dillaman) > * librbd: fix merge-diff for >2GB diff-files (#14030, Jason Dillaman) > * librbd: invalidate object map on error even w/o holding lock (#1337= 2, Jason Dillaman) > * librbd: reads larger than cache size hang (#13164, Lu Shi) > * mds: ceph mds add_data_pool check for EC pool is wrong (#12426, Joh= n Spray) > * mon: MonitorDBStore: get_next_key() only if prefix matches (#11786,= Joao Eduardo Luis) > * mon: OSDMonitor: do not assume a session exists in send_incremental= () (#14236, Joao Eduardo Luis) > * mon: check for store writeablility before participating in election= (#13089, Sage Weil) > * mon: compact full epochs also (#14537, Kefu Chai) > * mon: include min_last_epoch_clean as part of PGMap::print_summary a= nd PGMap::dump (#13198, Guang Yang) > * mon: map_cache can become inaccurate if osd does not receive the os= dmaps (#10930, Kefu Chai) > * mon: should not set isvalid =3D true when cephx_verify_authorizer r= eturn false (#13525, Ruifeng Yang) > * osd: Ceph Pools' MAX AVAIL is 0 if some OSDs' weight is 0 (#13840, = Chengyuan Li) > * osd: FileStore calls syncfs(2) even it is not supported (#12512, Ke= fu Chai) > * osd: FileStore: potential memory leak if getattrs fails. (#13597, x= ie xingguo) > * osd: IO error on kvm/rbd with an erasure coded pool tier (#12012, K= efu Chai) > * osd: OSD::build_past_intervals_parallel() shall reset primary and u= p_primary when begin a new past_interval. (#13471, xiexingguo) > * osd: ReplicatedBackend: populate recovery_info.size for clone (bug = symptom is size mismatch on replicated backend on a clone in scrub) (#1= 2828, Samuel Just) > * osd: ReplicatedPG: wrong result code checking logic during sparse_r= ead (#14151, xie xingguo) > * osd: ReplicatedPG::hit_set_trim osd/ReplicatedPG.cc: 11006: FAILED = assert(obc) (#13192, #9732, #12968, Kefu Chai, Zhiqiang Wang, Samuel Ju= st, David Zafman) > * osd: avoid multi set osd_op.outdata in tier pool (#12540, Xinze Chi= ) > * osd: bug with cache/tiering and snapshot reads (#12748, Kefu Chai) > * osd: ceph osd pool stats broken in hammer (#13843, BJ Lougee) > * osd: ceph-disk prepare fails if device is a symlink (#13438, Joe Ju= lian) > * osd: check for full before changing the cached obc (hammer) (#13098= , Alexey Sheplyakov) > * osd: config_opts: increase suicide timeout to 300 to match recovery= (#14376, Samuel Just) > * osd: disable filestore_xfs_extsize by default (#14397, Ken Dreyer) > * osd: do not cache unused memory in attrs (#12565, Xinze Chi, Ning Y= ao) > * osd: dumpling incrementals do not work properly on hammer and newer= (#13234, Samuel Just) > * osd: filestore: fix peek_queue for OpSequencer (#13209, Xinze Chi) > * osd: hit set clear repops fired in same epoch as map change -- segf= ault since they fall into the new interval even though the repops are c= leared (#12809, Samuel Just) > * osd: object_info_t::decode() has wrong version (#13462, David Zafma= n) > * osd: osd/OSD.cc: 2469: FAILED assert(pg_stat_queue.empty()) on shut= down (#14212, Sage Weil) > * osd: osd/PG.cc: 288: FAILED assert(info.last_epoch_started >=3D inf= o.history.last_epoch_started) (#14015, David Zafman) > * osd: osd/PG.cc: 3837: FAILED assert(0 =3D=3D "Running incompatible = OSD") (#11661, David Zafman) > * osd: osd/ReplicatedPG: Recency fix (#14320, Sage Weil, Robert LeBla= nc) > * osd: pg stuck in replay (#13116, Sage Weil) > * osd: race condition detected during send_failures (#13821, Sage Wei= l) > * osd: randomize scrub times (#10973, Kefu Chai) > * osd: requeue_scrub when kick_object_context_blocked (#12515, Xinze = Chi) > * osd: revert: use GMT time for hitsets (#13812, Loic Dachary) > * osd: segfault in agent_work (#13199, Samuel Just) > * osd: should recalc the min_last_epoch_clean when decode PGMap (#131= 12, Kefu Chai) > * osd: smaller object_info_t xattrs (#14803, Sage Weil) > * osd: we do not ignore notify from down osds (#12990, Samuel Just) > * rbd: QEMU hangs after creating snapshot and stopping VM (#13726, Ja= son Dillaman) > * rbd: TaskFinisher::cancel should remove event from SafeTimer (#1447= 6, Douglas Fuller) > * rbd: avoid re-writing old-format image header on resize (#13674, Ja= son Dillaman) > * rbd: fix bench-write (#14225, Sage Weil) > * rbd: rbd-replay does not check for EOF and goes to endless loop (#1= 4452, Mykola Golub) > * rbd: rbd-replay-prep and rbd-replay improvements (#13221, #13220, #= 13378, Jason Dillaman) > * rbd: verify self-managed snapshot functionality on image create (#1= 3633, Jason Dillaman) > * rgw: Make RGW_MAX_PUT_SIZE configurable (#6999, Vladislav Odintsov,= Yuan Zhou) > * rgw: Setting ACL on Object removes ETag (#12955, Brian Felton) > * rgw: backport content-type casing (#12939, Robin H. Johnson) > * rgw: bucket listing hangs on versioned buckets (#12913, Yehuda Sade= h) > * rgw: fix wrong etag calculation during POST on S3 bucket. (#11241, = Vladislav Odintsov, Radoslaw Zarzynski) > * rgw: get bucket location returns region name, not region api name (= #13458, Yehuda Sadeh) > * rgw: missing handling of encoding-type=3Durl when listing keys in b= ucket (#12735, Jeff Weber) > * rgw: orphan tool should be careful about removing head objects (#12= 958, Yehuda Sadeh) > * rgw: orphans finish segfaults (#13824, Igor Fedotov) > * rgw: rgw-admin: document orphans commands in usage (#14516, Yehuda = Sadeh) > * rgw: swift API returns more than real object count and bytes used w= hen retrieving account metadata (#13140, Sangdi Xu) > * rgw: swift use Civetweb ssl can not get right url (#13628, Weijun D= uan) > * rgw: value of Swift API's X-Object-Manifest header is not url_decod= ed during segment look up (#12728, Radoslaw Zarzynski) > * tests: fixed broken Makefiles after integration of ttng into rados = (#13210, Sebastien Ponce) > * tests: fsx failed to compile (#14384, Greg Farnum) > * tests: notification slave needs to wait for master (#13810, Jason D= illaman) > * tests: qa: remove legacy OS support from rbd/qemu-iotests (#13483, = #14385, Vasu Kulkarni, Jason Dillaman) > * tests: testprofile must be removed before it is re-created (#13664,= Loic Dachary) > * tools: ceph-monstore-tool must do out_store.close() (#10093, huangj= un) > * tools: heavy memory shuffling in rados bench (#12946, Piotr Da=C5=82= ek) > * tools: race condition in rados bench (#12947, Piotr Da=C5=82ek) > * tools: tool for artificially inflate the leveldb of the mon store f= or testing purposes (#10093, #11815, #14217, Cilang Zhao, Bo Cai, Kefu= Chai, huangjun, Joao Eduardo Luis) > > Getting Ceph > ------------ > > * Git at git://github.com/ceph/ceph.git > * Tarball at http://download.ceph.com/tarballs/ceph-0.94.6.tar.gz > * For packages, see http://ceph.com/docs/master/install/get-packages > * For ceph-deploy, see http://ceph.com/docs/master/install/install-ce= ph-deploy -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html