1 commit e832001feaf8c176593e0325c8298e3f16dfb403 (tag: refs/tags/v0.94.6, refs/remotes/gh/hammer)
2 Author: Jenkins Build Slave User <jenkins-build@jenkins-slave-wheezy.localdomain>
3 Date: Mon Feb 22 21:10:17 2016 +0000
7 commit 7abb6ae8f3cba67009bd022aaeee0a87cdfc6477
9 Author: Loic Dachary <loic@dachary.org>
10 Date: Mon Feb 8 14:19:57 2016 +0700
12 Merge pull request #7501 from SUSE/wip-14624-hammer
14 hammer: fsx failed to compile
16 Reviewed-by: Loic Dachary <ldachary@redhat.com>
18 commit ceb6fcc5f2828995bfe59bfc4e206010a4fe3230
19 Merge: 2f4e1d1 f2ca42b
20 Author: Sage Weil <sage@redhat.com>
21 Date: Fri Feb 5 21:10:46 2016 -0500
23 Merge pull request #7524 from ktdreyer/wip-14637-hammer-man-radosgw-admin-orphans
25 hammer: doc: regenerate man pages, add orphans commands to radosgw-admin(8)
27 commit 2f4e1d1ff8e91fc2ee9c23d5a17c3174d15a7103
28 Merge: 2ca3c3e 9ab5fd9
29 Author: Orit Wasserman <owasserm@redhat.com>
30 Date: Fri Feb 5 10:30:22 2016 +0100
32 Merge pull request #7526 from ceph/wip-14516-hammer
34 hammer: rgw-admin: document orphans commands in usage
36 commit 2ca3c3e5683ef97902d0969e49980d69c81b4034
37 Merge: 02353f6 5c8d1d7
38 Author: Loic Dachary <loic@dachary.org>
39 Date: Fri Feb 5 12:47:33 2016 +0700
41 Merge pull request #7441 from odivlad/backport-pr-14569
43 [backport] hammer: rgw: Make RGW_MAX_PUT_SIZE configurable
45 Reviewed-by: Yehuda Sadeh <ysadehwe@redhat.com>
47 commit 02353f6baa460949d7dd2738346d8d748401bc5b
48 Merge: f3bab8c 0e1378e
49 Author: Loic Dachary <loic@dachary.org>
50 Date: Fri Feb 5 12:46:54 2016 +0700
52 Merge pull request #7442 from odivlad/backport-pr-14570
54 [backport] rgw: fix wrong etag calculation during POST on S3 bucket.
56 Reviewed-by: Yehuda Sadeh <ysadehwe@redhat.com>
58 commit 9ab5fd9d67da43e986489e4f580a597dd8cb551e
59 Author: Yehuda Sadeh <yehuda@redhat.com>
60 Date: Mon Feb 1 16:33:55 2016 -0800
62 rgw-admin: document orphans commands in usage
66 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
67 (cherry picked from commit 105a76bf542e05b739d5a03ca8ae55432350f107)
70 src/rgw/rgw_admin.cc (trivial resolution)
72 commit 0e1378effdd1d0d70d3de05c79b208e9f8b8e328
73 Author: Radoslaw Zarzynski <rzarzynski@mirantis.com>
74 Date: Mon Sep 21 20:32:29 2015 +0200
76 [backport] rgw: fix wrong etag calculation during POST on S3 bucket.
79 (cherry picked from commit 742906a)
81 Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
82 Signed-off-by: Vladislav Odintsov <odivlad@gmail.com>
83 Tested-by: Vladislav Odintsov <odivlad@gmail.com>
85 commit 5c8d1d74069f70b85bc4286e6d1136bce1dc593f
86 Author: Yuan Zhou <yuan.zhou@intel.com>
87 Date: Thu Jul 9 16:56:07 2015 +0800
89 [backport] rgw: Make RGW_MAX_PUT_SIZE configurable
91 The 5GB limit of a single operation uploading was part of S3 spec.
92 However some private setups may have some special requirements
93 on this limit. It's more convinent to have a configurable value.
95 Closes: http://tracker.ceph.com/issues/14569
96 (cherry picked from commit df97f28)
98 Signed-off-by: Yuan Zhou <yuan.zhou@intel.com>
99 Signed-off-by: Vladislav Odintsov <odivlad@gmail.com>
100 Tested-by: Vladislav Odintsov <odivlad@gmail.com>
102 commit f2ca42b1218182f4bfa27718c9606705d8b9941f
103 Author: Ken Dreyer <kdreyer@redhat.com>
104 Date: Wed Feb 3 19:51:58 2016 -0700
106 doc: add orphans commands to radosgw-admin(8)
108 The --help text was added in 105a76bf542e05b739d5a03ca8ae55432350f107.
109 Add the corresponding entries to the man page.
113 Signed-off-by: Ken Dreyer <kdreyer@redhat.com>
115 (cherry picked from commit ec162f068b40f594c321df5caa9fe2541551b89e)
116 Cherry-pick to hammer includes nroff source change (in master the
117 nroff sources are no longer present in Git.)
119 commit e42ed6d6414ad55d671dd0f406b1dababd643af8
120 Author: Ken Dreyer <kdreyer@redhat.com>
121 Date: Thu Feb 4 11:04:39 2016 -0700
123 man: rebuild manpages
125 following the procedure in admin/manpage-howto.txt.
127 Signed-off-by: Ken Dreyer <kdreyer@redhat.com>
129 commit a8fc6a9ffb2f8021657ad412fd0aaaaf7f98bd53
130 Author: Greg Farnum <gfarnum@redhat.com>
131 Date: Wed Jan 13 13:17:53 2016 -0800
133 fsx: checkout old version until it compiles properly on miras
135 I sent a patch to xfstests upstream at
136 http://article.gmane.org/gmane.comp.file-systems.fstests/1665, but
137 until that's fixed we need a version that works in our test lab.
139 Signed-off-by: Greg Farnum <gfarnum@redhat.com>
140 (cherry picked from commit 7d52372ae74878ebd001036ff0a7aad525eb15b6)
142 commit f3bab8c7bc6bba3c79329d7106684596634e17a6
143 Merge: 31d86b1 1b02859
144 Author: Loic Dachary <loic@dachary.org>
145 Date: Wed Feb 3 12:41:56 2016 +0700
147 Merge pull request #7454 from dachary/wip-14584-hammer
149 hammer: fsstress.sh fails
151 Reviewed-by: Loic Dachary <ldachary@redhat.com>
153 commit 31d86b1580d59581d130e7a5e84905a5b4d67666
154 Merge: 2c8e579 2817ffc
155 Author: Loic Dachary <loic@dachary.org>
156 Date: Wed Feb 3 11:38:57 2016 +0700
158 Merge pull request #6918 from asheplyakov/hammer-bug-12449
160 osd: check for full before changing the cached obc (hammer)
162 Reviewed-by: Loic Dachary <ldachary@redhat.com>
164 commit 1b02859480677abbd8708650764ed6815917e0cd
165 Author: Yan, Zheng <zyan@redhat.com>
166 Date: Tue Aug 18 15:22:55 2015 +0800
168 qa/fsstress.sh: fix 'cp not writing through dangling symlink'
170 On some test machines, /usr/lib/ltp/testcases/bin/fsstress is
171 dangling symlink. 'cp -f' is impotent in this case.
174 Signed-off-by: Yan, Zheng <zyan@redhat.com>
175 (cherry picked from commit 479f2a760baf6af4714d3391a366a6b3acf1bcaf)
177 commit 2c8e57934284dae0ae92d1aa0839a87092ec7c51
178 Merge: 1cab151 700be56
179 Author: Sage Weil <sage@redhat.com>
180 Date: Sat Jan 30 21:42:29 2016 -0500
182 Merge pull request #7236 from athanatos/wip-14376
184 config_opts: increase suicide timeout to 300 to match recovery
186 http://pulpito.ceph.com/sage-2016-01-30_09:58:32-rados-wip-sage-testing-hammer---basic-mira/
188 Reviewed-by: Sage Weil <sage@redhat.com>
190 commit 1cab151a6cdc8a1b9b2f11518b77ea149c19d067
191 Merge: a134c44 5105d50
192 Author: Sage Weil <sage@redhat.com>
193 Date: Sat Jan 30 21:42:12 2016 -0500
195 Merge pull request #6450 from dachary/wip-13672-hammer
197 tests: testprofile must be removed before it is re-created
199 http://pulpito.ceph.com/sage-2016-01-30_09:58:32-rados-wip-sage-testing-hammer---basic-mira/
201 Reviewed-by: Sage Weil <sage@redhat.com>
203 commit a134c44d626ead3308464474d67604363bac6f5d
204 Merge: 520792d 3f1292e
205 Author: Sage Weil <sage@redhat.com>
206 Date: Sat Jan 30 21:41:39 2016 -0500
208 Merge pull request #6680 from SUSE/wip-13859-hammer
210 hammer: ceph.spec.in License line does not reflect COPYING
212 http://pulpito.ceph.com/sage-2016-01-30_09:58:32-rados-wip-sage-testing-hammer---basic-mira/
214 commit 520792d51e949c451767d116e267d86fee812ada
215 Merge: 4d0fafb c2c6d02
216 Author: Sage Weil <sage@redhat.com>
217 Date: Sat Jan 30 21:41:18 2016 -0500
219 Merge pull request #6791 from branch-predictor/bp-5812-backport
221 Race condition in rados bench
223 http://pulpito.ceph.com/sage-2016-01-30_09:58:32-rados-wip-sage-testing-hammer---basic-mira/
225 commit 4d0fafb289fc35f44e6e74bb974c402ba147d4d4
226 Merge: 211a093 6379ff1
227 Author: Sage Weil <sage@redhat.com>
228 Date: Sat Jan 30 21:40:38 2016 -0500
230 Merge pull request #6973 from dreamhost/wip-configure-hammer
232 configure.ac: no use to add "+" before ac_ext=c
234 http://pulpito.ceph.com/sage-2016-01-30_09:58:32-rados-wip-sage-testing-hammer---basic-mira/
236 Reviewed-by: Kefu Chai <kchai@redhat.com>
238 commit 211a093d4107c1806ad7e4876bc5c550a5b5b7d4
239 Merge: 0c13656 53742bd
240 Author: Sage Weil <sage@redhat.com>
241 Date: Sat Jan 30 21:40:13 2016 -0500
243 Merge pull request #7206 from dzafman/wip-14292
245 osd/PG.cc: 3837: FAILED assert(0 == "Running incompatible OSD")
247 http://pulpito.ceph.com/sage-2016-01-30_09:58:32-rados-wip-sage-testing-hammer---basic-mira/
249 Reviewed-by: David Zafman <dzafman@redhat.com>
251 commit 0c136561600e295ec48dcf29a77aa2cd293a7236
252 Merge: 1ea14ba ae56de0
253 Author: Sage Weil <sage@redhat.com>
254 Date: Sat Jan 30 21:39:42 2016 -0500
256 Merge pull request #7207 from rldleblanc/recency_fix_for_hammer
258 hammer: osd/ReplicatedPG: Recency fix for Hammer
260 Reviewed-by: Sage Weil <sage@redhat.com>
262 commit 1ea14bae10930ed3e66b7b0140e359009a44275e
263 Merge: 1740d8c 8d9e08c
264 Author: Sage Weil <sage@redhat.com>
265 Date: Sat Jan 30 21:39:11 2016 -0500
267 Merge pull request #7347 from tchaikov/wip-hammer-10093
269 tools: ceph-monstore-tool must do out_store.close()
271 http://pulpito.ceph.com/sage-2016-01-30_09:58:32-rados-wip-sage-testing-hammer---basic-mira/
273 Reviewed-by: Sage Weil <sage@redhat.com>
275 commit 1740d8ce8eb7e7debce1684a19736783489628c3
276 Merge: 7848cdc c7252a3
277 Author: Sage Weil <sage@redhat.com>
278 Date: Sat Jan 30 21:38:35 2016 -0500
280 Merge pull request #7411 from dachary/wip-14467-hammer
282 hammer: disable filestore_xfs_extsize by default
284 http://pulpito.ceph.com/sage-2016-01-30_09:58:32-rados-wip-sage-testing-hammer---basic-mira/
286 Reviewed-by: Sage Weil <sage@redhat.com>
288 commit 7848cdc4c8c3598cd291d26a2dd4d76abc4bcda9
289 Merge: 57abeab 70f1ba3
290 Author: Sage Weil <sage@redhat.com>
291 Date: Sat Jan 30 21:38:13 2016 -0500
293 Merge pull request #7412 from dachary/wip-14470-hammer
295 tools: tool for artificially inflate the leveldb of the mon store for testing purposes
297 Reviewed-by: Sage Weil <sage@redhat.com>
299 commit 57abeab1f2610d89f0ab2a028c8e093afde5e403
300 Merge: 4d7d7c3 9109304
301 Author: Sage Weil <sage@redhat.com>
302 Date: Sat Jan 30 21:37:46 2016 -0500
304 Merge pull request #7446 from liewegas/wip-14537-hammer
306 mon: compact full epochs also
308 http://pulpito.ceph.com/sage-2016-01-30_09:58:32-rados-wip-sage-testing-hammer---basic-mira/
310 Reviewed-by: Sage Weil <sage@redhat.com>
312 commit 4d7d7c33ff58ffeb27b47d379bf7dd4af90785b0
313 Merge: 8360486 8c28f2f
314 Author: Josh Durgin <jdurgin@redhat.com>
315 Date: Sat Jan 30 11:45:31 2016 -0800
317 Merge pull request #7182 from dachary/wip-14143-hammer
319 hammer: Verify self-managed snapshot functionality on image create
321 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
323 commit 8360486764c3ab4e10a34f5bdf7555e4c3385977
324 Merge: 501e01a b2961ce
325 Author: Josh Durgin <jdurgin@redhat.com>
326 Date: Sat Jan 30 11:45:20 2016 -0800
328 Merge pull request #7183 from dachary/wip-14283-hammer
330 hammer: rbd: fix bench-write
332 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
334 commit 501e01acc55f846cd3ad85e8294a9598c1c90907
335 Merge: 97d4f6a 24c0b27
336 Author: Josh Durgin <jdurgin@redhat.com>
337 Date: Sat Jan 30 11:45:05 2016 -0800
339 Merge pull request #7416 from dachary/wip-14466-hammer
341 hammer: rbd-replay does not check for EOF and goes to endless loop
343 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
345 commit 97d4f6a7bed5a0d47f18b3a223a916aef355eaf9
346 Merge: 51cc015 46d626d
347 Author: Josh Durgin <jdurgin@redhat.com>
348 Date: Sat Jan 30 11:44:50 2016 -0800
350 Merge pull request #7417 from dachary/wip-14553-hammer
352 hammer: rbd: TaskFinisher::cancel should remove event from SafeTimer
354 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
356 commit 51cc015c7e20d2ea9b9517238481fb80687be17c
357 Merge: 23c24fc 73e03de
358 Author: Josh Durgin <jdurgin@redhat.com>
359 Date: Sat Jan 30 11:44:32 2016 -0800
361 Merge pull request #7407 from dillaman/wip-14543-hammer
363 librbd: ImageWatcher shouldn't block the notification thread
365 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
367 commit 23c24fc50d07a07db12239ac372bfce5a9fe2255
368 Merge: 9513391 26e832e
369 Author: Josh Durgin <jdurgin@redhat.com>
370 Date: Sat Jan 30 11:44:12 2016 -0800
372 Merge pull request #6980 from dillaman/wip-14063-hammer
374 librbd: fix merge-diff for >2GB diff-files
376 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
378 commit 91093041a86fda40de08a366b5118e5e3ae275f0 (refs/remotes/me/wip-14537-hammer)
379 Author: Kefu Chai <kchai@redhat.com>
380 Date: Thu Jan 28 02:09:53 2016 -0800
382 mon: compact full epochs also
384 by compacting the ${prefix}.${start}..${prefix}..${end} does not
385 necessary compact the range of ${prefix}."full_"${start}..
386 ${prefix}."full_"${end}. so when more and more epochs get trimmed
387 with out a full range compaction, the size of monitor store could
391 Signed-off-by: Kefu Chai <kchai@redhat.com>
392 (cherry picked from commit 93d633a25ad8c36c972bb766c38187e2612041e1)
394 commit 2817ffcf4e57f92551b86388681fc0fe70c386ec
395 Author: Alexey Sheplyakov <asheplyakov@mirantis.com>
396 Date: Wed Dec 16 15:31:52 2015 +0300
398 Check for full before changing the cached obc
400 ReplicatedPG::prepare_transaction(): check if the pool is full before
401 updating the cached ObjectContext to avoid the discrepancy between
402 the cached and the actual object size (and other metadata).
403 While at it improve the check itself: consider cluster full flag,
404 not just the pool full flag, also consider object count changes too,
407 Based on commit a1eb380c3d5254f9f1fe34b4629e51d77fe010c1
411 Signed-off-by: Alexey Sheplyakov <asheplyakov@mirantis.com>
413 commit 951339103d35bc8ee2de880f77aada40d15b592a (refs/remotes/gh/wip-test-14716-4)
414 Merge: e43aca5 5e5b512
415 Author: Loic Dachary <loic@dachary.org>
416 Date: Fri Jan 29 23:31:47 2016 +0700
418 Merge pull request #6353 from theanalyst/wip-13513-hammer
420 rgw: value of Swift API's X-Object-Manifest header is not url_decoded during segment look up
422 Reviewed-by: Yehuda Sadeh <ysadehwe@redhat.com>
424 commit e43aca51d630274a57237b0f91a56df79ce8874a
425 Merge: 7e20e6e 4420929
426 Author: Loic Dachary <loic@dachary.org>
427 Date: Fri Jan 29 23:31:16 2016 +0700
429 Merge pull request #6620 from SUSE/wip-13820-hammer
431 hammer: rgw: Setting ACL on Object removes ETag
433 Reviewed-by: Yehuda Sadeh <ysadehwe@redhat.com>
435 commit 7e20e6e9d6626c5ac8b1f479011ab77a9e87da23
436 Merge: f1d5570 cbb5c1f
437 Author: Loic Dachary <loic@dachary.org>
438 Date: Fri Jan 29 23:30:57 2016 +0700
440 Merge pull request #7186 from dachary/wip-13888-hammer
442 hammer: rgw: orphans finish segfaults
444 Reviewed-by: Yehuda Sadeh <ysadehwe@redhat.com>
446 commit ae56de09fe1385958b5b600d1f0c91383989926f
447 Author: Sage Weil <sage@redhat.com>
448 Date: Wed Nov 25 14:40:26 2015 -0500
450 osd: recency should look at newest (not oldest) hitsets
452 Reported-by: xinxin shu <xinxin.shu@intel.com>
453 Signed-off-by: Sage Weil <sage@redhat.com>
454 (cherry picked from commit 528eae92d010ea34ae8cf0e8b2290aaa5e058d24)
457 src/osd/ReplicatedPG.cc
458 code section was moved to ReplicatedPG::maybe_promote
460 Signed-off-by: Robert LeBlanc <robert.leblanc@endurance.com>
462 commit 5cefcb975771f0c2efb7dfc77ce14a93a4ee7f1b
463 Author: Sage Weil <sage@redhat.com>
464 Date: Wed Nov 25 14:39:08 2015 -0500
466 osd/ReplicatedPG: fix promotion recency logic
468 Recency is defined as how many of the last N hitsets an object
469 must appear in in order to be promoted. The previous logic did
470 nothing of the sort... it checked for the object in any one of
471 the last N hitsets, which led to way to many promotions and killed
472 any chance of the cache performing properly.
474 While we are here, we can simplify the code to drop the max_in_*
475 fields (no longer necessary).
477 Note that we may still want a notion of 'temperature' that does
478 tolerate the object missing in one of the recent hitsets.. but
479 that would be different than recency, and should probably be
480 modeled after the eviction temperature model.
482 Backport: infernalis, hammer
483 Reported-by: Nick Fisk <nick@fisk.me.uk>
484 Signed-off-by: Sage Weil <sage@redhat.com>
485 (cherry picked from commit 180c8743addc5ae2f1db9c58cd2996ca6e7ac18b)
488 src/osd/ReplicatedPG.cc
489 code section was moved to ReplicatedPG::maybe_promote
491 Signed-off-by: Robert LeBlanc <robert.leblanc@endurance.com>
493 commit f1d5570beab0769b925b917e402d441ff053794c
494 Merge: c4bb343 50c82f2
495 Author: Sage Weil <sage@redhat.com>
496 Date: Fri Jan 29 08:52:51 2016 -0500
498 Merge pull request #5789 from SUSE/wip-12928-hammer
500 rpm: libcephfs_jni1 has no %post and %postun
502 Reviewed-by: Sage Weil <sage@redhat.com>
504 commit c4bb34320df8bfa734512a400fe8664c131363ff
505 Merge: 86ba6ca a5e4f70
506 Author: Sage Weil <sage@redhat.com>
507 Date: Fri Jan 29 08:50:56 2016 -0500
509 Merge pull request #7434 from tchaikov/wip-14441-hammer
511 man: document listwatchers cmd in "rados" manpage
513 Reviewed-by: Sage Weil <sage@redhat.com>
515 commit a5e4f70af422b05ece53f245bc15491bb1dd540d
516 Author: Kefu Chai <kchai@redhat.com>
517 Date: Wed Dec 23 11:23:38 2015 +0800
519 man: document listwatchers cmd in "rados" manpage
521 Signed-off-by: Kefu Chai <kchai@redhat.com>
522 (cherry picked from commit c2e391301efc43f0b431e89737246b2c43bf10a9)
524 man/rados.8: also added the rendered man.8 man page, as we don't
525 put the generated man pages in master anymore. but
526 they are still in the hammer's source repo.
528 commit 46d626d92d7af218816d831cfcc1265b3ea31b72
529 Author: Douglas Fuller <dfuller@redhat.com>
530 Date: Fri Jan 22 11:18:40 2016 -0800
532 rbd: remove canceled tasks from timer thread
534 When canceling scheduled tasks using the timer thread, TaskFinisher::cancel
535 does not call SafeTimer::cancel_event, so events fire anyway. Add this call.
538 Signed-off-by: Douglas Fuller <dfuller@redhat.com>
539 (cherry picked from commit 2aa0f318c862dbe3027d74d345671506605778eb)
541 commit 24c0b27c6f6a26c2b7bab5bcbc421a18592d026f
542 Author: Mykola Golub <mgolub@mirantis.com>
543 Date: Thu Jan 21 13:45:42 2016 +0200
545 rbd-replay: handle EOF gracefully
548 Signed-off-by: Mykola Golub <mgolub@mirantis.com>
549 (cherry picked from commit c59b84c3e2c9bbda68219e4d2288a889dd9ca6cb)
552 src/rbd_replay/BufferReader.cc: trivial resolution
553 caused by s/CEPH_PAGE_SIZE/CEPH_BUFFER_APPEND_SIZE/
554 in the context of one hunk.
556 commit 70f1ba33fe1a81d2631d54429749433a6cbfca44
557 Author: Cilang Zhao <zhao.cilang@h3c.com>
558 Date: Tue Jan 5 14:34:05 2016 +0800
560 tools: monstore: add 'show-versions' command.
562 Using this tool, the first/last committed version of maps will be shown.
564 Signed-off-by: Cilang Zhao <zhao.cilang@h3c.com>
565 (cherry picked from commit 21e6ba0c18428caff45733e6b43d197be38af8bb)
567 commit 926017187910c9e6a3fb8babf9b498cf07941819
568 Author: Kefu Chai <kchai@redhat.com>
569 Date: Wed Sep 16 18:28:52 2015 +0800
571 tools: ceph_monstore_tool: add inflate-pgmap command
573 this command repeatly add the latest pgmap to the monstore in order
574 to inflate it. the command helps with the testing of some monstore
575 related performance issue of monitor
578 Signed-off-by: Kefu Chai <kchai@redhat.com>
579 (cherry picked from commit b0f6af814a58dd374ca95e84a4e381f8ef1368de)
581 commit a1d5728c928eb7e6b8c655741a0db3398ced7d56
582 Author: Bo Cai <cai.bo@h3c.com>
583 Date: Tue Oct 20 15:23:49 2015 +0800
585 tools:support printing the crushmap in readable fashion.
587 Signed-off-by: Bo Cai <cai.bo@h3c.com>
588 (cherry picked from commit b7faf678ed0241abc7eb86b9baaa6db669a22d53)
590 commit bd9484750f546974047dacd9176a8072be846596
591 Author: Bo Cai <cai.bo@h3c.com>
592 Date: Mon Sep 14 19:50:47 2015 +0800
594 tools:print the map infomation in human readable format.
596 Signed-off-by: Bo Cai <cai.bo@h3c.com>
597 (cherry picked from commit 5ed8cdc19150382c946a373ec940d76f98e6ecb7)
599 commit fba65832aad8a46d94a9256a56997e9df9e62297
600 Author: Bo Cai <cai.bo@h3c.com>
601 Date: Mon Sep 14 19:19:05 2015 +0800
603 tools:remove the local file when get map failed.
605 Signed-off-by: Bo Cai <cai.bo@h3c.com>
606 (cherry picked from commit 0b03b32d8ba76fe9f6f1158e68eb440e3670393a)
608 commit 1bb899a290b77188b44a53ef7c7a40910c9248b2
609 Author: Joao Eduardo Luis <joao@suse.de>
610 Date: Mon Jul 13 12:35:13 2015 +0100
612 tools: ceph_monstore_tool: describe behavior of rewrite command
614 Signed-off-by: Joao Eduardo Luis <joao@suse.de>
615 (cherry picked from commit a881f9385feb0f5a61fa22357984d6f291c08177)
617 commit 9035c69481f4aa4786414ac41cbc36fb4a3ca51d
618 Author: Kefu Chai <kchai@redhat.com>
619 Date: Fri Jun 19 22:57:57 2015 +0800
621 tools/ceph-monstore-tools: add rewrite command
623 "rewrite" command will
624 - add a new osdmap version to update current osdmap held by OSDMonitor
625 - add a new paxos version, as a proposal it will
626 * rewrite all osdmap epochs from specified epoch to the last_committed
627 one with the specified crush map.
628 * add the new osdmap which is added just now
629 so the leader monitor can trigger a recovery process to apply the transaction
630 to all monitors in quorum, and hence bring them back to normal after being
631 injected with a faulty crushmap.
634 Signed-off-by: Kefu Chai <kchai@redhat.com>
635 (cherry picked from commit 306373427836ca0c2418dbe6caab26d74d94d12e)
637 commit 90aaed744243dfc7c620f91e19fc0bfa233f711f
638 Author: huangjun <hjwsm1989@gmail.com>
639 Date: Tue Apr 21 14:04:40 2015 +0800
641 tools: ceph-monstore-tool must do out_store.close()
643 this change fixes the "store-copy" command.
644 Like the bug reported in http://tracker.ceph.com/issues/10093.
646 Signed-off-by: huangjun <hjwsm1989@gmail.com>
647 (cherry picked from commit d85e0f8c50fce62be012506f7ffcb1cdeb0c819c)
649 commit c7252a3d54bae88f91c2b4e63fc9c27cfbb2423e
650 Author: Ken Dreyer <kdreyer@redhat.com>
651 Date: Mon Jan 18 08:24:46 2016 -0700
653 osd: disable filestore_xfs_extsize by default
655 This option involves a tradeoff: When disabled, fragmentation is worse,
656 but large sequential writes are faster. When enabled, large sequential
657 writes are slower, but fragmentation is reduced.
660 (cherry picked from commit aed85775bf53c273786ce4999320134822722af5)
662 commit 86ba6caf992d2544cdd174e3b3f26a6099c91fc5
663 Merge: b6b8ee4 0325f8a
664 Author: Loic Dachary <loic@dachary.org>
665 Date: Fri Jan 29 10:36:05 2016 +0700
667 Merge pull request #7316 from ceph/wip-deb-lttng-hammer
669 deb: strip tracepoint libraries from Wheezy/Precise builds
671 All other "modern" Debian-based OSes have a functional LTTng-UST. Since only hammer needs to build on these older distros, this fix only affects the deb building process for those two releases(since autoconf detects that LTTng is broken).
673 Reviewed-by: Loic Dachary <ldachary@redhat.com>
675 commit b6b8ee41fc77679e82304e79da6dbd0c35f5c067
676 Merge: d54840b f96c812
677 Author: Loic Dachary <loic@dachary.org>
678 Date: Fri Jan 29 10:21:43 2016 +0700
680 Merge pull request #7187 from dachary/wip-13831-hammer
682 hammer: init script reload doesn't work on EL7
684 Reviewed-by: Loic Dachary <ldachary@redhat.com>
686 commit 73e03def9271fb5d1739b195e428c3ebfcebd59b
687 Author: Jason Dillaman <dillaman@redhat.com>
688 Date: Thu Jan 28 14:38:20 2016 -0500
690 librbd: ImageWatcher shouldn't block the notification thread
692 Blocking the notification thread will also result in librados async
693 callbacks becoming blocked (since they use the same thread).
695 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
696 (cherry picked from commit 7e2019a72733dff43e55c9b22df12939d584f87d)
699 src/librbd/ImageWatcher.[cc|h]: fewer RPC messages
701 commit 5617166f78c1995436b4e0794dab2d8254331815
702 Author: Jason Dillaman <dillaman@redhat.com>
703 Date: Thu Jan 28 14:35:54 2016 -0500
705 librados_test_stub: watch/notify now behaves similar to librados
707 Notifications are executed via the same librados AIO callback
708 thread, so it's now possible to catch deadlock.
710 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
711 (cherry picked from commit 43722571838a2d78ca8583b5a1ea381cd988de0b)
713 commit 8fc82b23304ef327933723373cd4d1090d04bfbc
714 Author: Jason Dillaman <dillaman@redhat.com>
715 Date: Thu Jan 28 12:40:18 2016 -0500
717 tests: simulate writeback flush during snap create
719 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
720 (cherry picked from commit e9570fcf4be9cd5c06937769d074fa52ffb32e4b)
722 commit d54840bf4a70fc65285bbfdff0c7bf8f579643b1
723 Merge: 4051bc2 9f30fe1
724 Author: Sage Weil <sage@redhat.com>
725 Date: Tue Jan 26 10:12:29 2016 -0500
727 Merge pull request #7365 from liewegas/wip-tcmalloc-hammer
729 osd: pass tcmalloc env through to ceph-osd
731 Reviewed-by: Ken Dreyer <kdreyer@redhat.com>
733 commit 9f30fe18e839f5846c90e3b5995171a0132d7f3a
734 Author: Sage Weil <sage@redhat.com>
735 Date: Tue Jan 26 08:43:15 2016 -0500
737 upstart/ceph-osd.conf: pass TCMALLOC_MAX_TOTAL_THREAD_CACHE_BYTES through
739 Signed-off-by: Sage Weil <sage@redhat.com>
740 (cherry picked from commit a95a3d34cccb9f9f0782a96ca05e5fe9c2b01772)
742 commit a58873e213d7423d89c95db4e1710dc9631e3313
743 Author: Sage Weil <sage@redhat.com>
744 Date: Wed Jan 20 18:36:08 2016 -0500
746 init-ceph: pass TCMALLOC_MAX_TOTAL_THREAD_CACHE_BYTES through
748 ..when set in the sysconfig/default file.
750 Signed-off-by: Sage Weil <sage@redhat.com>
751 (cherry picked from commit 03e01ae263ef207046933890821ae75d5e60d9b8)
753 [source default/sysconfig file too]
755 commit 8d9e08c44150a4f3ad06ec1abc130b9aeaf479d9
756 Author: huangjun <hjwsm1989@gmail.com>
757 Date: Tue Apr 21 14:04:40 2015 +0800
759 tools: ceph-monstore-tool must do out_store.close()
761 this change fixes the "store-copy" command.
762 Like the bug reported in http://tracker.ceph.com/issues/10093.
764 Signed-off-by: huangjun <hjwsm1989@gmail.com>
765 (cherry picked from commit d85e0f8c50fce62be012506f7ffcb1cdeb0c819c)
767 commit 0325f8af5cbee3d74e9f363f61c2e2ababf501d9
768 Author: Jason Dillaman <dillaman@redhat.com>
769 Date: Thu Jan 21 15:36:37 2016 -0500
771 deb: strip tracepoint libraries from Wheezy/Precise builds
773 These releases do not (by default) have a sane LTTng-UST environment,
774 which results in autoconf disabling support for tracing.
776 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
778 commit 4051bc2a5e4313ac0f6236d7a34ed5fb4a1d9ea2
779 Merge: 4e67418 e4d3e9b
780 Author: Jason Dillaman <dillaman@redhat.com>
781 Date: Mon Jan 18 20:46:07 2016 -0500
783 Merge pull request #7252 from ceph/wip-13483-hammer
785 qa: remove legacy OS support from rbd/qemu-iotests
787 Reviewed-by: Jason Dillaman <dillaman@redhat.com>
789 commit e4d3e9b29aa6a5f9efff8b787949a5bbf48ad7de (refs/remotes/gh/wip-13483-hammer)
790 Author: Jason Dillaman <dillaman@redhat.com>
791 Date: Mon Jan 18 15:57:43 2016 -0500
793 qa: disable rbd/qemu-iotests test case 055 on RHEL/CentOS
796 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
798 commit f7acd44c26310242b69ee50322bd6b43fdc774b9
799 Author: Jason Dillaman <dillaman@redhat.com>
800 Date: Thu Oct 15 12:44:06 2015 -0400
802 qa: remove legacy OS support from rbd/qemu-iotests
804 RHEL7 derivatives were failing test 002 since they were using
805 legacy test cases for now unsupported OSes.
808 Signed-off-by: Vasu Kulkarni <vasu@redhat.com>
810 commit 8c28f2f28d960d823ffd632671edaf029c30fb0f
811 Author: Jason Dillaman <dillaman@redhat.com>
812 Date: Mon Dec 14 17:41:49 2015 -0500
814 librbd: optionally validate RBD pool configuration (snapshot support)
817 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
818 (cherry picked from commit 1fea4dadc60e13518e9ee55d136fbc4e9d3a621e)
821 src/common/config_opts.h: trivial resolution
823 commit 700be56c530879a72a628c62265d18f0a5d8fb3b
824 Author: Samuel Just <sjust@redhat.com>
825 Date: Thu Jan 14 08:35:23 2016 -0800
827 config_opts: increase suicide timeout to 300 to match recovery
830 Backport: hammer, firefly
831 Signed-off-by: Samuel Just <sjust@redhat.com>
833 commit 4e67418958e5caf5e4f81c4ed566e8c7269930fa
834 Merge: 28e99a9 1ab2b48
835 Author: Sage Weil <sage@redhat.com>
836 Date: Thu Jan 14 09:21:50 2016 -0500
838 Merge pull request #7179 from dachary/wip-14287-hammer
840 hammer: ReplicatedPG: wrong result code checking logic during sparse_read
842 Reviewed-by: Sage Weil <sage@redhat.com>
844 commit 28e99a90b38ce8c0285c1aec9c5524365385be00
845 Merge: 17a2965 86f5cf6
846 Author: Sage Weil <sage@redhat.com>
847 Date: Thu Jan 14 09:20:53 2016 -0500
849 Merge pull request #7178 from dachary/wip-14285-hammer
851 hammer: osd/OSD.cc: 2469: FAILED assert(pg_stat_queue.empty()) on shutdown
853 Reviewed-by: Sage Weil <sage@redhat.com>
855 commit 17a2965825494765249c244aef59ebd821711d42
856 Merge: a1459ea b0856ee
857 Author: Sage Weil <sage@redhat.com>
858 Date: Thu Jan 14 09:20:34 2016 -0500
860 Merge pull request #7177 from dachary/wip-14043-hammer
862 hammer: osd/PG.cc: 288: FAILED assert(info.last_epoch_started >= info.history.last_epoch_started)
864 Reviewed-by: Sage Weil <sage@redhat.com>
866 commit a1459eacd7cc0d064493d01f963b8c1231cd3518
867 Merge: c9a13a1 9cee89b
868 Author: Sage Weil <sage@redhat.com>
869 Date: Thu Jan 14 09:20:10 2016 -0500
871 Merge pull request #7180 from dachary/wip-14288-hammer
873 hammer: ceph osd pool stats broken in hammer
875 Reviewed-by: Sage Weil <sage@redhat.com>
877 commit c9a13a12e88c9a18bfb3cc99d4a7c103b4330426
878 Merge: 174a1a9 4d0b9a1
879 Author: Sage Weil <sage@redhat.com>
880 Date: Thu Jan 14 09:19:18 2016 -0500
882 Merge pull request #6994 from badone/wip-13993-hammer
884 log: Log.cc: Assign LOG_DEBUG priority to syslog calls
886 Reviewed-by: Sage Weil <sage@redhat.com>
888 commit 174a1a974725ce4025be4dfdf3b13df766dfac76
889 Merge: cb167e9 6cf261c
890 Author: Sage Weil <sage@redhat.com>
891 Date: Thu Jan 14 09:18:55 2016 -0500
893 Merge pull request #6839 from SUSE/wip-13789-hammer
895 Objecter: potential null pointer access when do pool_snap_list.
897 Reviewed-by: Sage Weil <sage@redhat.com>
899 commit cb167e998bbbdf1b9c273f6bc4f118796d396d37
900 Merge: 1c4ba85 66ff0aa
901 Author: Sage Weil <sage@redhat.com>
902 Date: Thu Jan 14 09:18:23 2016 -0500
904 Merge pull request #6835 from SUSE/wip-13892-hammer
906 hammer: auth/cephx: large amounts of log are produced by osd
908 Reviewed-by: Sage Weil <sage@redhat.com>
910 commit 1c4ba855b34290069c7ce0fa990fe72b7a1b381e
911 Merge: 72b6b68 61da13b
912 Author: Sage Weil <sage@redhat.com>
913 Date: Thu Jan 14 09:17:53 2016 -0500
915 Merge pull request #6834 from SUSE/wip-13930-hammer
917 hammer: Ceph Pools' MAX AVAIL is 0 if some OSDs' weight is 0
919 Reviewed-by: Sage Weil <sage@redhat.com>
921 commit 72b6b68f43ffc953ca7f3056574bd68428a5abe8
922 Merge: d064f9f 53e81aa
923 Author: Sage Weil <sage@redhat.com>
924 Date: Thu Jan 14 09:17:25 2016 -0500
926 Merge pull request #6832 from SUSE/wip-13936-hammer
928 hammer: Ceph daemon failed to start, because the service name was already used.
930 Reviewed-by: Sage Weil <sage@redhat.com>
932 commit d064f9f6b3bc93fa5887304a9b9ab92d503ebd15
933 Merge: 0ca6285 ac05617
934 Author: Sage Weil <sage@redhat.com>
935 Date: Thu Jan 14 09:16:54 2016 -0500
937 Merge pull request #6755 from SUSE/wip-13870-hammer
939 hammer: OSD: race condition detected during send_failures
941 Reviewed-by: Sage Weil <sage@redhat.com>
943 commit 0ca62855f22136c7042a5988366d46e59629f206
944 Merge: e9f545b a81bcf7
945 Author: Loic Dachary <loic@dachary.org>
946 Date: Thu Jan 14 08:22:35 2016 +0100
948 Merge pull request #6415 from dillaman/wip-13541-hammer
950 LTTng-UST tracing should be dynamically enabled
952 Reviewed-by: Loic Dachary <ldachary@redhat.com>
954 commit e9f545b4a44c91faf91a8ae3824b1b0d46909244
955 Merge: 8521916 b2f1e76
956 Author: Loic Dachary <loic@dachary.org>
957 Date: Thu Jan 14 08:15:02 2016 +0100
959 Merge pull request #7176 from dachary/wip-13440-hammer
961 ceph-disk prepare fails if device is a symlink
963 Reviewed-by: Loic Dachary <ldachary@redhat.com>
965 commit 85219163f9b78faa3c1fe96cad7129c6813550c0
966 Merge: ebfb3fb 5264bc6
967 Author: Loic Dachary <loic@dachary.org>
968 Date: Thu Jan 14 08:14:31 2016 +0100
970 Merge pull request #7150 from jecluis/wip-14236
972 mon: OSDMonitor: do not assume a session exists in send_incremental()
974 Reviewed-by: Kefu Chai <kchai@redhat.com>
976 commit ebfb3fbe8a1920474941d4a95540cc4a53a747d4
977 Merge: e1b9208 447ab1d
978 Author: Josh Durgin <jdurgin@redhat.com>
979 Date: Wed Jan 13 17:15:32 2016 -0800
981 Merge pull request #7226 from dillaman/wip-13810-hammer
983 tests: notification slave needs to wait for master
985 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
987 commit 447ab1d70a63fceed39518d5714c138c54f6525e
988 Author: Jason Dillaman <dillaman@redhat.com>
989 Date: Wed Jan 13 12:44:01 2016 -0500
991 tests: notification slave needs to wait for master
993 If the slave instance starts before the master, race
994 conditions are possible.
997 Backport: infernalis, hammer
998 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
999 (cherry picked from commit 3992d6fe67bbf82322cedc1582406caaf6d4de60)
1001 commit 53742bdbc587747c93413f9db38f3eeb71487872
1002 Author: David Zafman <dzafman@redhat.com>
1003 Date: Mon Jun 15 17:55:41 2015 -0700
1005 ceph_osd: Add required feature bits related to this branch to osd_required mask
1007 Signed-off-by: David Zafman <dzafman@redhat.com>
1008 (cherry picked from commit 0b2bab460c013ea5cece10ea028d26da3485eaf5)
1010 commit 3066231865e2fe56344de9db26024ac65e03053d
1011 Author: David Zafman <dzafman@redhat.com>
1012 Date: Thu Jun 4 18:47:42 2015 -0700
1014 osd: CEPH_FEATURE_CHUNKY_SCRUB feature now required
1016 Feature present since at least the Dumpling release.
1017 A later commit will add it to the osd_required mask
1021 Signed-off-by: David Zafman <dzafman@redhat.com>
1022 (cherry picked from commit 7a10a7e196efd0f59fa7adb87d0a120655b227d8)
1024 commit e1b92081c9e4b21eb30cc873c239083a08fce12f
1025 Merge: 9708e59 3d3595f
1026 Author: Loic Dachary <loic@dachary.org>
1027 Date: Tue Jan 12 14:28:26 2016 +0100
1029 Merge pull request #6530 from SUSE/wip-13760-hammer
1031 unknown argument --quiet in udevadm settle
1033 Reviewed-by: Loic Dachary <ldachary@redhat.com>
1035 commit 6379ff16dbc7fdecccaa8874d978d1ab58ce44cf
1036 Author: Kefu Chai <kchai@redhat.com>
1037 Date: Tue May 5 15:07:33 2015 +0800
1039 configure.ac: no use to add "+" before ac_ext=c
1041 Signed-off-by: Kefu Chai <kchai@redhat.com>
1042 (cherry picked from commit 899dd23766c5ae6fef39bf24ef0692127d17deb3)
1043 Signed-off-by: Robin H. Johnson <robin.johnson@dreamhost.com>
1045 commit f96c8125d423f24e52df0f8420b0b2748e34ca90
1046 Author: Herve Rousseau <hroussea@cern.ch>
1047 Date: Fri Nov 6 09:52:28 2015 +0100
1049 rgw: fix reload on non Debian systems.
1051 When using reload in non-debian systems, /bin/sh's kill is used to send the HUP signal to the radosgw process.
1052 This kill version doesn't understand -SIGHUP as a valid signal, using -HUP does work.
1056 Signed-off-by: Hervé Rousseau <hroussea@cern.ch>
1057 (cherry picked from commit 1b000abac3a02d1e788bf25eead4b6873133f5d2)
1059 commit cbb5c1fc2386205277ad22474c1f696a07fcb972
1060 Author: Igor Fedotov <ifedotov@mirantis.com>
1061 Date: Thu Nov 19 13:38:40 2015 +0300
1063 Fixing NULL pointer dereference
1065 Signed-off-by: Igor Fedotov <ifedotov@mirantis.com>
1066 (cherry picked from commit 93d3dfe0441be50a6990d458ee0ee3289af39b20)
1068 commit b2961cef3d8e84d2302815e33eb7dc9033d2bb78
1069 Author: Sage Weil <sage@redhat.com>
1070 Date: Tue Aug 18 16:05:29 2015 -0400
1072 rbd: fix bench-write
1074 Make each IO get a unique offset!
1076 Signed-off-by: Sage Weil <sage@redhat.com>
1077 (cherry picked from commit 333f3a01a9916c781f266078391c580efb81a0fc)
1079 commit 9cee89bbdd1d89d2fe72c8d73fc0c18db515198a
1080 Author: BJ Lougee <almightybeeij@gmail.com>
1081 Date: Mon Apr 13 11:24:38 2015 -0500
1083 Check that delta_sum.stats.sum.num_object_copies and delta_sum.stats.sum.num_object are greater than zero
1085 This fixes division by zero.
1087 Signed-off-by BJ Lougee <almightybeeij@gmail.com>
1089 (cherry picked from commit 27ed729c1088133400aa072eeca9e125942f2d94)
1091 commit 1ab2b481596ad5296e168bf8027e59d41190176e
1092 Author: xiexingguo <258156334@qq.com>
1093 Date: Tue Dec 22 17:05:06 2015 +0800
1095 ReplicatedPG: fix sparse-read result code checking logic
1097 Move ahead the result code checking logic before we continue to verify the trailing hole, otherwise
1098 the real result of non-hole reading may be overwritten and thus confuse caller.
1101 Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
1102 (cherry picked from commit b3aa29e324cf0a96a1f2f5dcf1ba998219457bcd)
1105 src/osd/ReplicatedPG.cc: trivial resolution
1107 commit 86f5cf6caa36760802775df19cdabe55bcafa33e
1108 Author: Sage Weil <sage@redhat.com>
1109 Date: Fri Jan 1 09:32:04 2016 -0500
1111 osd: clear pg_stat_queue after stopping pgs
1114 Signed-off-by: Sage Weil <sage@redhat.com>
1115 (cherry picked from commit c25ff9927febe77cda31b534971b62f75c196ca2)
1117 commit b0856ee67c0e7cf6ab6095d3f657c18014859526
1118 Author: David Zafman <dzafman@redhat.com>
1119 Date: Thu Dec 3 14:52:24 2015 -0800
1121 osd: Test osd_find_best_info_ignore_history_les config in another assert
1123 Signed-off-by: David Zafman <dzafman@redhat.com>
1124 (cherry picked from commit 02a9a41f151a3d968bf8066749658659dc6e3ac4)
1126 commit b2f1e76d8e95b6e2f88fa1c122ea8cd24532cd10
1127 Author: Joe Julian <jjulian@io.com>
1128 Date: Fri Oct 9 12:57:06 2015 -0700
1130 Compare parted output with the dereferenced path
1132 Compare parted output with the dereferenced path of the device as parted
1133 prints that instead of the symlink we called it with.
1135 http://tracker.ceph.com/issues/13438 Fixes: #13438
1137 Signed-off-by: Joe Julian <jjulian@io.com>
1138 (cherry picked from commit b3c7cb098195111b9c642e5a9b726b63717f2e0d)
1140 commit 9708e59a1fc2ae52cfae848ce585751bc9fbe572
1141 Merge: 9739d4d b62cac6
1142 Author: Loic Dachary <loic@dachary.org>
1143 Date: Mon Jan 11 09:01:12 2016 +0100
1145 Merge pull request #6512 from SUSE/wip-13734-hammer
1147 rgw: swift API returns more than real object count and bytes used when retrieving account metadata
1149 Reviewed-by: Yehuda Sadeh <ysadehwe@redhat.com>
1151 commit 5264bc677f0f612165987bf51fe7d7b4af32fa77
1152 Author: Joao Eduardo Luis <joao@suse.de>
1153 Date: Thu Jan 7 19:20:47 2016 +0000
1155 mon: OSDMonitor: do not assume a session exists in send_incremental()
1157 We may not have an open session for a given osd. If we blindly assume we
1158 do, we may end up trying to send incrementals we do not have to the osd.
1160 And then we will crash.
1162 This fixes a regression introduced by
1164 171fee1b82d2675e364da7f96dfb9dd286d9b6e6
1166 which is meant as a backport of
1168 de43a02e06650a552f048dc8acd17f255126fed9
1170 but so happens to intruduce a line that wasn't on the original patch. We
1171 imagine it was meant to make the 's->osd_epoch' assignment work without
1172 checking the session, as per the original patch, but the backporter must
1173 have forgotten to also backport the assertion on the not-null session.
1174 The unfortunate introduction of the check for a not-null session
1175 triggered this regression.
1177 The regression itself is due to enforcing that a session exists for the
1178 osd we are sending the incrementals to. However, if we come via the
1179 OSDMonitor::process_failures() path, that may very well not be the case,
1180 as we are handling potentially-old MOSDFailure messages that may no
1181 longer have an associated session. By enforcing the not-null session, we
1182 don't check whether we have the requested versions (i.e., if
1183 our_earliest_version <= requested_version), and thus we end up on the
1184 path that assumes that we DO HAVE all the necessary versions -- when we
1185 may not, thus finally asserting because we are reading blank
1186 incremental versions.
1190 Signed-off-by: Joao Eduardo Luis <joao@suse.de>
1192 commit 4d0b9a1bbbabe7b27279a7b6e0a45f5b0d920c66
1193 Author: Brad Hubbard <bhubbard@redhat.com>
1194 Date: Mon Dec 7 11:31:28 2015 +1000
1196 log: Log.cc: Assign LOG_DEBUG priority to syslog calls
1199 Signed-off-by: Brad Hubbard <bhubbard@redhat.com>
1200 (cherry picked from commit 8e93f3f45db681f82633ca695a7dc4e7bd030584)
1202 commit 26e832e76de90e0a751868b044ea745a97a5af82
1203 Author: Jason Dillaman <dillaman@redhat.com>
1204 Date: Fri Dec 18 15:22:13 2015 -0500
1206 librbd: fix merge-diff for >2GB diff-files
1209 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
1211 commit 9739d4de49f8167866eda556b2f1581c068ec8a7
1212 Merge: d064636 a9d3f07
1213 Author: Sage Weil <sage@redhat.com>
1214 Date: Thu Dec 17 10:16:29 2015 -0500
1216 Merge pull request #6544 from liewegas/wip-smaller-object-info
1218 osd: make encoded object_info_t smaller to fit inside the XFS inode
1220 Reviewed-by: Samuel Just <sjust@redhat.com>
1222 commit a9d3f078cb6cdaf7f8778ba1a07d333548a9d1a2
1223 Author: Sage Weil <sage@redhat.com>
1224 Date: Mon Dec 14 13:13:33 2015 -0500
1226 osd/osd_types: skip encoding newer object_info_t fields if they are unused
1228 This reduces the size of the encoded object_info_t in most cases,
1229 enough to get us under the 255 byte limit for a single inline
1232 Signed-off-by: Sage Weil <sage@redhat.com>
1234 commit 1548a3fda7dd63e28792140e7e8ad1ac9b706e49
1235 Author: Sage Weil <sage@redhat.com>
1236 Date: Mon Dec 14 13:00:27 2015 -0500
1238 osd/ReplicatedPG: do not set local_mtime on non-tiered pool
1240 If a pool isn't tiered, don't bother with setting local_mtime. The only
1241 users are the tiering agent (which isn't needed if there is not tiering)
1242 and scrub for deciding if an object should get its digest recorded (we can
1245 Signed-off-by: Sage Weil <sage@redhat.com>
1247 commit 98bdb09bb8d632b2df0b252b3d9676fd9511a1c8
1248 Author: Sage Weil <sage@redhat.com>
1249 Date: Mon Dec 14 12:59:29 2015 -0500
1251 osd/PGBackend: use mtime for digest decision if local_mtime is empty
1253 If we don't have a local_mtime value, use mtime instead, for the purposes
1254 of deciding if we should record a digest after scrub.
1256 Signed-off-by: Sage Weil <sage@redhat.com>
1258 commit fe1c28dea4e5607a0c72eab1f046074616cd55a7
1259 Author: Sage Weil <sage@redhat.com>
1260 Date: Mon Dec 14 11:35:46 2015 -0500
1262 osd/ReplicatedPG: do not set omap digest if there is no omap
1264 We want to avoid encoding it if we can. And if the FLAG_OMAP is not set
1265 we don't need to *also* store an empty crc.
1267 Signed-off-by: Sage Weil <sage@redhat.com>
1269 commit 6cf261cd2e80adbaf2898cb354c8159a57749e9d
1270 Author: xiexingguo <258156334@qq.com>
1271 Date: Mon Nov 2 21:46:11 2015 +0800
1273 Objecter: remove redundant result-check of _calc_target in _map_session.
1275 Result-code check is currently redundant since _calc_target never returns a negative value.
1276 Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
1278 (cherry picked from commit 5a6117e667024f51e65847f73f7589467b6cb762)
1280 commit 5d0b5c1389d8c0baddb40cc5ff8f1d9cc6a4d315
1281 Author: xiexingguo <258156334@qq.com>
1282 Date: Thu Oct 29 17:32:50 2015 +0800
1284 Objecter: potential null pointer access when do pool_snap_list.
1286 Objecter: potential null pointer access when do pool_snap_list. Shall check pool existence first.
1288 Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
1289 (cherry picked from commit 865541605b6c32f03e188ec33d079b44be42fa4a)
1291 commit 66ff0aa0130973aa2bea3a1a4ec8f4e3e6c19da8
1292 Author: qiankunzheng <zheng.qiankun@h3c.com>
1293 Date: Thu Nov 5 07:29:49 2015 -0500
1295 auth/cephx: large amounts of log are produced by osd
1296 if the auth of osd is deleted when the osd is running, the osd will produce large amounts of log.
1299 Signed-off-by: Qiankun Zheng <zheng.qiankun@h3c.com>
1300 (cherry picked from commit 102f0b19326836e3b0754b4d32da89eb2bc0b03c)
1302 commit 61da13bbe271964c56116fb2bed51045290f951a
1303 Author: Chengyuan Li <chengyli@ebay.com>
1304 Date: Thu Nov 19 22:29:39 2015 -0700
1306 mon/PGMonitor: MAX AVAIL is 0 if some OSDs' weight is 0
1308 In get_rule_avail(), even p->second is 0, it's possible to be used
1309 as divisor and quotient is infinity, then is converted to an integer
1310 which is negative value.
1311 So we should check p->second value before calculation.
1313 It fixes BUG #13840.
1315 Signed-off-by: Chengyuan Li <chengyli@ebay.com>
1316 (cherry picked from commit 18713e60edd1fe16ab571f7c83e6de026db483ca)
1318 commit 53e81aab442855b34ee6e922f181bae3bf33e292
1319 Author: wangchaunhong <root@A22832429.(none)>
1320 Date: Tue Oct 20 18:40:23 2015 +0800
1322 init-ceph: fix systemd-run cant't start ceph daemon sometimes
1325 Signed-off-by: Chuanhong Wang <wang.chuanhong@zte.com.cn>
1326 (cherry picked from commit 2f36909e1e08bac993e77d1781a777b386335669)
1330 different content of cmd variable
1332 commit c2c6d02591519dfd15ddcb397ac440322a964deb
1333 Author: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
1334 Date: Mon Jul 6 09:56:11 2015 +0200
1336 tools: fix race condition in seq/rand bench
1338 Under certain conditions (like bench seq/rand -b 1024 -t 128) it is
1339 possible that aio_read reads data into destination buffers before or
1340 during memcmp execution, resulting in "[..] is not correct!" errors
1341 even if actual objects are perfectly fine.
1342 Also, moved latencty calculation around, so it is no longer affeted
1345 Signed-off-by: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
1348 src/common/obj_bencher.cc
1350 commit a619b621b0a7c670eeaf163d9e2b742d13c9f517
1351 Author: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
1352 Date: Wed May 20 12:41:22 2015 +0200
1354 tools: add --no-verify option to rados bench
1356 When doing seq and rand read benchmarks using rados bench, a quite large
1357 portion of cpu time is consumed by doing object verification. This patch
1358 adds an option to disable this verification when it's not needed, in turn
1359 giving better cluster utilization. rados -p storage bench 600 rand scores
1360 without --no-verification:
1362 Total time run: 600.228901
1363 Total reads made: 144982
1365 Bandwidth (MB/sec): 966
1370 Average Latency: 0.0662
1378 Same command, but with --no-verify:
1380 Total time run: 600.161379
1381 Total reads made: 174142
1383 Bandwidth (MB/sec): 1.16e+03
1388 Average Latency: 0.0551
1390 Min latency: 0.00343
1396 Note the decreased latencies, increased bandwidth and more reads performed.
1398 Signed-off-by: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
1399 (cherry picked from commit ca6abca63de813c83a6960f83624be8e1a86a1f8)
1402 src/common/obj_bencher.cc
1403 src/common/obj_bencher.h
1405 commit d06463604cb8daeda288e824e8812352c0d6a7d9
1406 Merge: ec35347 609f256
1407 Author: Loic Dachary <loic@dachary.org>
1408 Date: Wed Dec 2 20:48:00 2015 +0100
1410 Merge pull request #6527 from theanalyst/wip-12856-hammer
1412 rgw: missing handling of encoding-type=url when listing keys in bucket
1414 Reviewed-by: Yehuda Sadeh <ysadehwe@redhat.com>
1416 commit ec35347d8fd7a0eb175710e6111a9196746e278c
1417 Merge: 487dc34 1b06f03
1418 Author: Loic Dachary <loic@dachary.org>
1419 Date: Wed Dec 2 20:47:43 2015 +0100
1421 Merge pull request #6491 from SUSE/wip-13716-hammer
1423 rgw:swift use Civetweb ssl can not get right url
1425 Reviewed-by: Yehuda Sadeh <ysadehwe@redhat.com>
1427 commit 487dc3436a8fabd1b1a31563071c907bdf18f525
1428 Merge: 18d3ba3 99b4d1d
1429 Author: Loic Dachary <loic@dachary.org>
1430 Date: Wed Dec 2 20:47:27 2015 +0100
1432 Merge pull request #6351 from theanalyst/wip-13538-hammer
1434 rgw: orphan tool should be careful about removing head objects
1436 Reviewed-by: Yehuda Sadeh <ysadehwe@redhat.com>
1438 commit 18d3ba30323a44c1cc45bf4e870b8d9aeb5c3b0b
1439 Merge: 36e4db3 6d89f4b
1440 Author: Loic Dachary <loic@dachary.org>
1441 Date: Wed Dec 2 20:47:09 2015 +0100
1443 Merge pull request #6349 from theanalyst/wip-13540-hammer
1445 rgw: get bucket location returns region name, not region api name
1447 Reviewed-by: Yehuda Sadeh <ysadehwe@redhat.com>
1449 commit 36e4db36f5ec9418672bbfe0dce6ae757f697efe
1450 Merge: 8c4145e db1cbe7
1451 Author: Loic Dachary <loic@dachary.org>
1452 Date: Wed Dec 2 20:46:49 2015 +0100
1454 Merge pull request #5910 from kmroz/wip-hammer-backport-content-type
1456 rgw: backport content-type casing
1458 Reviewed-by: Yehuda Sadeh <ysadehwe@redhat.com>
1460 commit ac05617b246d3c2a329b1b99b0371e3f6b8541e9
1461 Author: Sage Weil <sage@redhat.com>
1462 Date: Thu Sep 17 21:42:53 2015 -0400
1464 osd: fix send_failures() locking
1466 It is unsafe to check failure_queue.empty() without the lock.
1468 Signed-off-by: Sage Weil <sage@redhat.com>
1469 (cherry picked from commit b3ca828ae8ebc9068073494c46faf3e8e1443ada)
1471 commit 8c4145ecc4a68accdb2120889fd933e8f6630dba
1472 Merge: 4804eec 112c686
1473 Author: Josh Durgin <jdurgin@redhat.com>
1474 Date: Wed Nov 25 08:47:10 2015 -0800
1476 Merge pull request #6587 from theanalyst/wip-13758-hammer
1478 common: pure virtual method called
1480 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
1482 commit 4804eec095959c3747b13d29224b2cfa5b7d198d
1483 Merge: 9b97e4f e693755
1484 Author: Daniel Gryniewicz <dang@fprintf.net>
1485 Date: Wed Nov 25 10:49:15 2015 -0500
1487 Merge pull request #6352 from theanalyst/wip-13536-hammer
1489 rgw: bucket listing hangs on versioned buckets
1491 commit 9b97e4f6a41bb3fe7bae92b71ae266361022cf5c
1492 Merge: 5a9e0a7 0378445
1493 Author: Loic Dachary <loic@dachary.org>
1494 Date: Wed Nov 25 16:48:09 2015 +0100
1496 Merge pull request #6589 from theanalyst/wip-13693-hammer
1498 osd: bug with cache/tiering and snapshot reads
1500 Reviewed-by: Kefu Chai <kchai@redhat.com>
1502 commit 5a9e0a722c654b83fc483e083ddb3035281f5397
1503 Merge: 3047f2b a322317
1504 Author: Loic Dachary <loic@dachary.org>
1505 Date: Wed Nov 25 16:22:10 2015 +0100
1507 Merge pull request #6585 from theanalyst/wip-13753-hammer
1509 rbd: avoid re-writing old-format image header on resize
1511 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
1513 commit 3047f2bc50d5d26e70e832bd427f15073bf7733c
1514 Merge: 407bd02 9c33dcc
1515 Author: Loic Dachary <loic@dachary.org>
1516 Date: Mon Nov 23 19:57:03 2015 +0100
1518 Merge pull request #6586 from theanalyst/wip-13755-hammer
1520 rbd: QEMU hangs after creating snapshot and stopping VM
1522 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
1524 commit 407bd0281582ebb3b50ce51bdb48244e22463c8f
1525 Merge: c3c400f bddbda1
1526 Author: Loic Dachary <loic@dachary.org>
1527 Date: Mon Nov 23 17:32:21 2015 +0100
1529 Merge pull request #6588 from theanalyst/wip-13770-hammer
1531 Objecter: pool op callback may hang forever.
1533 Reviewed-by: Sage Weil <sage@redhat.com>
1535 commit c3c400f68d9c2593acd9485d1214d14af4e930d0
1536 Merge: d116959 a52f7cb
1537 Author: Loic Dachary <loic@dachary.org>
1538 Date: Mon Nov 23 15:13:24 2015 +0100
1540 Merge pull request #6430 from SUSE/wip-13654-hammer
1542 crush: crash if we see CRUSH_ITEM_NONE in early rule step
1544 Reviewed-by: Sage Weil <sage@redhat.com>
1546 commit 3f1292e117ab1cdcaef3b24ee33854f7be142795
1547 Author: Nathan Cutler <ncutler@suse.com>
1548 Date: Thu Sep 3 20:30:50 2015 +0200
1550 ceph.spec.in: fix License line
1552 This is closer to my reading of
1553 https://github.com/ceph/ceph/blob/master/COPYING than the previous version.
1555 http://tracker.ceph.com/issues/12935 Fixes: #12935
1557 Signed-off-by: Nathan Cutler <ncutler@suse.com>
1558 (cherry picked from commit ef7418421b3748c712019c8aedd02b8005c1e1ea)
1560 commit d116959442f67c8f36898ca989b490ca84a609c6
1561 Merge: 12be099 74203b8
1562 Author: Loic Dachary <loic@dachary.org>
1563 Date: Mon Nov 23 09:28:09 2015 +0100
1565 Merge pull request #6420 from SUSE/wip-13637-hammer
1567 FileStore: potential memory leak if getattrs fails.
1569 Reviewed-by: Kefu Chai <kchai@redhat.com>
1571 commit 12be099a7b0f2692c167af96928e939b2227160c
1572 Merge: 39c70d4 2052187
1573 Author: Loic Dachary <loic@dachary.org>
1574 Date: Sat Nov 21 09:10:27 2015 +0100
1576 Merge pull request #6528 from theanalyst/wip-13695-hammer
1578 init-rbdmap uses distro-specific functions
1580 Reviewed-by: Boris Ranto <branto@redhat.com>
1582 commit 39c70d4364a1be39f7c393847417f44279b4364c
1583 Merge: 65aeba0 9643ee6
1584 Author: Loic Dachary <loic@dachary.org>
1585 Date: Sat Nov 21 09:07:13 2015 +0100
1587 Merge pull request #6499 from SUSE/wip-13692-hammer
1589 osd: do not cache unused memory in attrs
1591 Reviewed-by: Samuel Just <sjust@redhat.com>
1593 commit 65aeba0b2f165d893d06df7e9bbe25f989edafe9
1594 Merge: 24d6698 394fbfc
1595 Author: Loic Dachary <loic@dachary.org>
1596 Date: Sat Nov 21 08:59:25 2015 +0100
1598 Merge pull request #6143 from dachary/wip-13340-hammer
1600 small probability sigabrt when setting rados_osd_op_timeout
1602 Reviewed-by: Sage Weil <sage@redhat.com>
1604 commit 24d66981d351848c44ae9e0395dad392f852be2b
1605 Merge: 5d204db 171fee1
1606 Author: Loic Dachary <loic@dachary.org>
1607 Date: Sat Nov 21 08:58:11 2015 +0100
1609 Merge pull request #5773 from tchaikov/wip-12835-hammer
1611 mon: map_cache can become inaccurate if osd does not receive the osdmaps
1613 Reviewed-by: Loic Dachary <ldachary@redhat.com>
1615 commit a81bcf723c1099f2bea5daf8b01b7d9853de323a
1616 Author: Jason Dillaman <dillaman@redhat.com>
1617 Date: Thu Oct 15 00:15:54 2015 -0400
1619 ceph_context: remove unsafe cast for singletons
1621 It was previously assumed that a CephContext singleton would
1622 inherit from CephContext::AssociatedSingletonObject, but it was
1623 not enforced. This could result in unknown behavior when the
1624 singleton is destroyed due to the implied virtual destructor.
1626 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
1627 (cherry picked from commit fb62c78637d7092f48871d943282f45029bd6d29)
1629 commit d50d7b2fa3751e8520694ee75eefe5ae56e57267
1630 Author: Jason Dillaman <dillaman@redhat.com>
1631 Date: Wed Sep 30 22:53:28 2015 -0400
1633 osd: conditionally initialize the tracepoint provider
1635 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
1636 (cherry picked from commit 0009f343a5daf28358a669836d95660c0d9068d8)
1638 commit bf34b3657339dc40c7939fcdddaf2b7ae78c82ad
1639 Author: Jason Dillaman <dillaman@redhat.com>
1640 Date: Wed Sep 30 22:37:05 2015 -0400
1642 librados: conditionally initialize the tracepoint provider
1644 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
1645 (cherry picked from commit 6368c281b53d6175a564725b85516de4b6ae54de)
1647 commit afc4f176f8371f0e61ce3463602f8a355b9283b9
1648 Author: Jason Dillaman <dillaman@redhat.com>
1649 Date: Wed Sep 30 22:10:10 2015 -0400
1651 librbd: conditionally initialize the tracepoint provider
1653 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
1654 (cherry picked from commit 1a6eea95bd1ccef53f01cf53e493e652faa203ee)
1657 src/librbd/librbd.cc: trivial resolution
1659 commit c82f93915875875b21ed779e7aaf297b8fbcdca1
1660 Author: Jason Dillaman <dillaman@redhat.com>
1661 Date: Wed Sep 30 19:17:08 2015 -0400
1663 common: new tracing config options
1665 Config options to enable LTTng-UST tracepoint providers for
1666 the OSD, OSD objectstore, librados, and librbd.
1668 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
1669 (cherry picked from commit 2ccef89badb1f7a63721a0bf48d05eb4c2d0a97a)
1672 src/common/config_opts.h: trivial resolution
1674 commit 2168c151066e0cea86284ffdf947a353f3b323d5
1675 Author: Jason Dillaman <dillaman@redhat.com>
1676 Date: Wed Sep 30 18:49:32 2015 -0400
1678 tracing: merge tracepoint provider makefiles into main makefile
1680 This results in the tracepoint provider shared libraries being
1681 placed in the library path for unittests.
1683 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
1684 (cherry picked from commit e11d8d68da27a1c224aa55fc40a72ac6aaa0441c)
1687 configure.ac: trivial resolution
1688 src/Makefile.am: trivial resolution
1690 commit d02beff1cf7650fe6e57cdafe64dcbee2631ed52
1691 Author: Jason Dillaman <dillaman@redhat.com>
1692 Date: Wed Sep 30 15:13:24 2015 -0400
1694 tracing: dynamic tracepoint provider helper
1696 The TracepointProvider class is a configuration observer. When
1697 tracing is enabled, it will dynamically load the associated
1698 tracepoint provider.
1700 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
1701 (cherry picked from commit b3d02cc21b4cc40a194f84a9bbbf82cf9e7956d1)
1704 src/common/Makefile.am: trivial resolution
1706 commit e53d66e42b1c3aea47832f7e8983284ec45d9efa
1707 Author: Jason Dillaman <dillaman@redhat.com>
1708 Date: Tue Sep 29 14:33:48 2015 -0400
1710 packaging: add new tracepoint probe shared libraries
1712 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
1713 (cherry picked from commit a7ed8e161b27c7852bf0f3a23a977ebd39137943)
1715 commit bb7c0f501918b75dee97052faeafb1d6db61bd0a
1716 Author: Jason Dillaman <dillaman@redhat.com>
1717 Date: Tue Sep 29 14:30:43 2015 -0400
1719 ceph.spec.in: add new tracepoint probe shared libraries
1721 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
1722 (cherry picked from commit f4feee2f9dcd7b94476701b66d1f0bdf6fb6e0c2)
1725 ceph.spec.in: trivial resolution
1727 commit e1da27134e4c9b4a2881aca664818598e5b2125b
1728 Author: Jason Dillaman <dillaman@redhat.com>
1729 Date: Tue Sep 29 14:13:46 2015 -0400
1731 lttng: move tracepoint probes to dynamic libraries
1733 LTTng-UST initializes itself at program load, which means it is
1734 currently always enabled. This can lead to issues with SElinux
1735 and AppArmor which might restrict access to the necessary device
1738 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
1739 (cherry picked from commit 4a5305e41677885d45b2b74299113a3d00189c66)
1742 src/test/Makefile-client.am: trivial resolution
1743 src/tracing/Makefile.am: trivial resolution
1745 commit 5d204db85b2ed84411cfabea9eb73aa04c6ce05a
1746 Merge: 1dbd4c6 8378aaf
1747 Author: Loic Dachary <loic@dachary.org>
1748 Date: Thu Nov 19 19:45:25 2015 +0100
1750 Merge pull request #6580 from dachary/wip-13786-hammer
1752 rbd-replay-* moved from ceph-test-dbg to ceph-common-dbg as well
1754 Reviewed-by: Loic Dachary <ldachary@redhat.com>
1756 commit 1dbd4c66307e615670c7569f107418876f8b977f
1757 Merge: 24a5dc0 634d7f6
1758 Author: Loic Dachary <loic@dachary.org>
1759 Date: Thu Nov 19 19:45:03 2015 +0100
1761 Merge pull request #6286 from dillaman/wip-13460-hammer
1763 rbd-replay-prep and rbd-replay improvements
1765 Reviewed-by: Loic Dachary <ldachary@redhat.com>
1767 commit 24a5dc0a6b4f73c92c82e1933a1924db3f8e1090
1768 Merge: d27da2e 6a40e4f
1769 Author: Loic Dachary <loic@dachary.org>
1770 Date: Thu Nov 19 19:41:20 2015 +0100
1772 Merge pull request #5790 from SUSE/wip-12932-hammer
1774 Miscellaneous spec file fixes
1776 Reviewed-by: Loic Dachary <ldachary@redhat.com>
1778 commit d27da2ee8fa3e074fdd91d414df2d59978db9497
1779 Merge: 7d21127 8358fb8
1780 Author: Kefu Chai <tchaikov@gmail.com>
1781 Date: Thu Nov 19 19:13:11 2015 +0800
1783 Merge pull request #6644 from dachary/wip-13812-upgrade
1785 revert: osd: use GMT time for hitsets
1787 Reviewed-by: Kefu Chai <kchai@redhat.com>
1789 commit 8358fb8946d8809d695092baa4a6abf5d5b5e265
1790 Author: Loic Dachary <loic@dachary.org>
1791 Date: Wed Nov 18 18:08:58 2015 +0100
1793 revert: osd: use GMT time for hitsets
1795 "Merge pull request #5825 from tchaikov/wip-12848-hammer"
1797 This reverts commit 39544718dc2f09bcfdc632ac72fd2a3cda87687e, reversing
1798 changes made to 4ad97162026e1eb6e6e948ddf3eb39f711431e45.
1800 http://tracker.ceph.com/issues/13812 Fixes: #13812
1802 Signed-off-by: Loic Dachary <loic@dachary.org>
1804 commit 4420929eacb67d17635fed53f15471a28942f31d
1805 Author: brian <bjfelton@gmail.com>
1806 Date: Wed Sep 23 09:49:36 2015 -0500
1808 rgw: fix modification to index attrs when setting acls
1812 - add ACL change after initial population
1813 - populate bufferlist with object attributes
1815 Signed-off-by: Brian Felton <bjfelton@gmail.com>
1816 (cherry picked from commit 7496741ebbd75b74d5ffeca5341cccb2318176e6)
1818 commit 7d21127f433afa2d9172954e7b8ff47c40d2d62b
1819 Merge: 6930601 1448915
1820 Author: Loic Dachary <loic@dachary.org>
1821 Date: Tue Nov 17 10:24:45 2015 +0100
1823 Merge pull request #6402 from SUSE/wip-13621-hammer
1825 CephFS restriction on removing cache tiers is overly strict
1827 Reviewed-by: Greg Farnum <gfarnum@redhat.com>
1829 commit 693060102594a2c505bed1e7e274731b554e9179
1830 Merge: c2daf09 3f33ce6
1831 Author: Loic Dachary <loic@dachary.org>
1832 Date: Mon Nov 16 22:06:51 2015 +0100
1834 Merge pull request #6354 from theanalyst/wip-13387-hammer
1836 librbd: reads larger than cache size hang
1838 Reviewed-by: Jason Dillaman <dillaman@redhat.com>
1840 commit c2daf09fc6b8802240e13367a0a2f3349a7cca56
1841 Merge: 1ca72fb af734e6
1842 Author: Loic Dachary <loic@dachary.org>
1843 Date: Mon Nov 16 19:48:07 2015 +0100
1845 Merge pull request #6289 from dillaman/wip-13461-hammer
1847 librbd: invalidate object map on error even w/o holding lock
1849 Reviewed-by: Loic Dachary <ldachary@redhat.com>
1851 commit 1ca72fbe6be0d664a3cbbd4356a8547c3b749a03
1852 Merge: ca573fe 3e65730
1853 Author: Loic Dachary <loic@dachary.org>
1854 Date: Mon Nov 16 18:06:25 2015 +0100
1856 Merge pull request #6153 from dachary/wip-13205-hammer
1858 ReplicatedBackend: populate recovery_info.size for clone (bug symptom is size mismatch on replicated backend on a clone in scrub)
1860 Reviewed-by: Samuel Just <sjust@redhat.com>
1862 commit ca573fe7ac5a699b77cacb8fedfa50fffc77dd52
1863 Merge: fdb3446 7161a2c
1864 Author: Loic Dachary <loic@dachary.org>
1865 Date: Mon Nov 16 18:06:12 2015 +0100
1867 Merge pull request #6158 from dachary/wip-13336-hammer
1869 osd: we do not ignore notify from down osds
1871 Reviewed-by: Samuel Just <sjust@redhat.com>
1873 commit fdb3446194743135b78a65c30a8285d001e563b6
1874 Merge: f017566 0ad9521
1875 Author: Loic Dachary <loic@dachary.org>
1876 Date: Mon Nov 16 18:06:01 2015 +0100
1878 Merge pull request #6335 from Abhishekvrshny/wip-13488-hammer
1880 object_info_t::decode() has wrong version
1882 Reviewed-by: Samuel Just <sjust@redhat.com>
1884 commit f017566823af70323b7d4cbe1f7b3b761378d07e
1885 Merge: a79acd4 9f3aebe
1886 Author: Loic Dachary <loic@dachary.org>
1887 Date: Mon Nov 16 18:05:39 2015 +0100
1889 Merge pull request #6401 from SUSE/wip-13620-hammer
1891 osd: pg stuck in replay
1893 Reviewed-by: Samuel Just <sjust@redhat.com>
1895 commit 8378aaf3f9d5463fb61a19d601201dd63a884419
1896 Author: Loic Dachary <ldachary@redhat.com>
1897 Date: Fri Nov 13 19:10:28 2015 +0100
1899 build/ops: rbd-replay moved from ceph-test-dbg to ceph-common-dbg
1901 http://tracker.ceph.com/issues/13785 Fixes: #13785
1903 Signed-off-by: Loic Dachary <loic@dachary.org>
1904 (cherry picked from commit a0204dc47acc0c50223e449fd23a2fc97bfd111a)
1907 debian/control: << 0.94.5-2 is appropriate for hammer
1909 commit 03784457ab77617ddafc048b01044858b1e65bd3
1910 Author: Kefu Chai <kchai@redhat.com>
1911 Date: Tue Sep 29 22:26:48 2015 +0800
1913 osd: fix the snapshot reads of evicted tiering pool
1915 reset ssc->exsits in finish_ctx() if the ctx->cache_evict is true, and
1916 the head is removed.
1919 Signed-off-by: Kefu Chai <kchai@redhat.com>
1920 (cherry picked from commit b61f3e43f1c9d43daa0dedd7cbd0fe5787cbdfbb)
1922 commit bddbda121c8d52ba8e50abd0341f14e332c6af5a
1923 Author: xiexingguo <258156334@qq.com>
1924 Date: Thu Oct 29 20:04:11 2015 +0800
1926 Objecter: pool_op callback may hang forever.
1928 pool_op callback may hang forever due to osdmap update during reply handling.
1930 Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
1932 (cherry picked from commit 00c6fa9e31975a935ed2bb33a099e2b4f02ad7f2)
1934 commit 112c686ffc98e3739c9944635f4044e2b34f210a
1935 Author: Jason Dillaman <dillaman@redhat.com>
1936 Date: Tue Jul 7 12:11:13 2015 -0400
1938 WorkQueue: new PointerWQ base class for ContextWQ
1940 The existing work queues do not properly function if added to a running
1941 thread pool. librbd uses a singleton thread pool which requires
1942 dynamically adding/removing work queues as images are opened and closed.
1945 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
1946 (cherry picked from commit 3e78b18b09d75626ca2599bac3b9f9c9889507a5)
1949 src/common/WorkQueue.h
1950 Trivial merge conflict at class `ContextWQ` initialization
1952 commit 9c33dccaad9a4cdd46e9ecfa1e3ba6c03d95885a
1953 Author: Jason Dillaman <dillaman@redhat.com>
1954 Date: Mon Nov 9 11:22:24 2015 -0500
1956 librbd: fixed deadlock while attempting to flush AIO requests
1958 In-flight AIO requests might force a flush if a snapshot was created
1959 out-of-band. The flush completion was previously invoked asynchronously,
1960 potentially via the same thread worker handling the AIO request. This
1961 resulted in the flush operation deadlocking since it can't complete.
1964 Backport: infernalis, hammer
1965 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
1966 (cherry picked from commit bfeb90e5fe24347648c72345881fd3d932243c98)
1968 commit b3b7877f9b4b3f43acab09d0dd6ee971b6aa1c29
1969 Author: Jason Dillaman <dillaman@redhat.com>
1970 Date: Mon Nov 9 10:48:10 2015 -0500
1972 tests: new test case to catch deadlock on RBD image refresh
1974 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
1975 (cherry picked from commit a9729d9553e7fb925509cad8d388cf52a9fede9c)
1977 commit a3223173f85c42147ff4ced730beffe85146a4ed
1978 Author: Jason Dillaman <dillaman@redhat.com>
1979 Date: Mon Nov 2 16:50:19 2015 -0500
1981 librbd: resize should only update image size within header
1983 Previously, the whole RBD image format 1 header struct was
1984 re-written to disk on a resize operation.
1987 Backport: infernalis, hammer, firefly
1988 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
1989 (cherry picked from commit d5be20b6d4646284571568ab28cbf45b0729390b)
1991 commit 6a40e4f19d7b4cd45a25161303c7363e96fe799e
1992 Author: Nathan Cutler <ncutler@suse.com>
1993 Date: Tue Sep 29 10:39:16 2015 +0200
1995 ceph.spec.in: lttng in SLES12 only
1997 Over in the SUSE sector, we are trying to enable the SLE_12 and openSUSE_13.2
1998 build targets. The lttng/babeltrace stuff is currently available only in
2001 Signed-off-by: Nathan Cutler <ncutler@suse.com>
2002 (cherry picked from commit bf9ca1eb107c4462d8768faf3ff4db7972384ffd)
2008 commit e508a44485366557ac8a280de35584f3b5edf720
2009 Author: Nathan Cutler <ncutler@suse.com>
2010 Date: Wed Jul 15 12:54:20 2015 +0200
2012 ceph.spec.in: fix lttng/babeltrace conditionals
2014 lttng and babeltrace are build dependencies for rbd-replay-prep. Make
2015 sure the right package names are used. Enable for SLE12, as well as
2016 for openSUSE 13.1 and higher.
2018 Move the BuildRequires out of the ceph-test subpackage and into the
2019 distro-conditional dependencies section.
2021 Make ordering of BuildRequires a little more alphabetical.
2023 http://tracker.ceph.com/issues/12360 Fixes: #12360
2025 Signed-off-by: Nathan Cutler <ncutler@suse.com>
2026 (cherry picked from commit e267128ce22d8b8cd83f6d6d82f24f496600e678)
2030 Dropped Requires: xmlstarlet in ceph-test that had inadvertently
2031 been grabbed from d2cc2b1
2033 commit 19c9546b79f506d0b4ee005a138a77f120c629e8
2034 Author: Ken Dreyer <kdreyer@redhat.com>
2035 Date: Tue Sep 8 10:41:02 2015 -0600
2037 packaging: move rbd-replay* to ceph-common
2039 The rbd-replay* utilities are useful for Ceph users with RBD clients.
2040 Currently the rbd-replay* utilities ship in the "ceph-test" package, and
2041 we intend this ceph-test package for Ceph developers and contributors,
2044 Move the rbd-replay* utilities to "ceph-common".
2046 http://tracker.ceph.com/issues/12994 Fixes: #12994
2048 Signed-off-by: Ken Dreyer <kdreyer@redhat.com>
2049 (cherry picked from commit 96486fd6f284ca3ab67f4f94631896d41c529e85)
2053 50a33dea has not been backported to hammer
2054 debian/ceph-test.install
2055 50a33dea has not been backported to hammer
2057 different ceph-test Replaces: and Breaks: version
2059 commit a79acd41187e6b049432bdc314f192e3fbb560a3
2060 Merge: 3d61493 31b7864
2061 Author: Loic Dachary <loic@dachary.org>
2062 Date: Wed Nov 11 11:59:36 2015 +0100
2064 Merge pull request #6213 from SUSE/wip-13425-hammer
2066 wrong conditional for boolean function KeyServer::get_auth()
2068 Reviewed-by: Kefu Chai <kchai@redhat.com>
2070 commit 3d61493337d481dec5139ccb6172710a118b2009
2071 Merge: 147f437 7ffd072
2072 Author: Loic Dachary <loic@dachary.org>
2073 Date: Wed Nov 11 06:59:55 2015 +0100
2075 Merge pull request #6336 from Abhishekvrshny/wip-13535-hammer
2077 LibRadosWatchNotify.WatchNotify2Timeout
2079 Reviewed-by: Sage Weil <sage@redhat.com>
2081 commit 147f437279e72c0766a126653e3db53abb73a4ac
2082 Merge: 1970e61 9085c82
2083 Author: Loic Dachary <loic@dachary.org>
2084 Date: Wed Nov 11 06:59:42 2015 +0100
2086 Merge pull request #6391 from SUSE/wip-13590-hammer
2088 mon: should not set isvalid = true when cephx_verify_authorizer return false
2090 Reviewed-by: Sage Weil <sage@redhat.com>
2092 commit 1970e618209a242fc0bd3c707dfdfb1a886d6a2c
2093 Merge: fb83704 ade0f1a
2094 Author: Loic Dachary <loic@dachary.org>
2095 Date: Wed Nov 11 06:58:55 2015 +0100
2097 Merge pull request #6398 from SUSE/wip-13588-hammer
2099 OSD::build_past_intervals_parallel() shall reset primary and up_primary when begin a new past_interval.
2101 Reviewed-by: Sage Weil <sage@redhat.com>
2103 commit 3d3595f86083a0f9847585eae4cffb8a82c816d4
2104 Author: Jason Dillaman <dillaman@redhat.com>
2105 Date: Tue Oct 27 10:13:27 2015 -0400
2107 krbd: remove deprecated --quiet param from udevadm
2109 This parameter has been removed since systemd 213, so this
2110 effects Fedora 21+, Debian Jessie, and potentially future
2114 Backport: hammer, infernalis
2115 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
2116 (cherry picked from commit 4300f2a9fe29627eea580564ff2d576de3647467)
2118 commit 4d81cd19087e049bf3c2fe0d10de9215852ab51d
2119 Author: Jason Dillaman <dillaman@redhat.com>
2120 Date: Tue Oct 27 10:12:34 2015 -0400
2122 run_cmd: close parent process console file descriptors
2124 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
2125 (cherry picked from commit f46f7dc94139c0bafe10361622416d7dc343d31f)
2127 commit fb83704d5db81aad95c61a61be52d07ce2d4d2b5
2128 Merge: 3954471 0742177
2129 Author: Loic Dachary <loic@dachary.org>
2130 Date: Tue Nov 10 20:24:35 2015 +0100
2132 Merge pull request #6199 from liewegas/wip-randomize-scrub-hammer
2134 osd: randomize scrub times
2136 Reviewed-by: Loic Dachary <ldachary@redhat.com>
2138 commit 39544718dc2f09bcfdc632ac72fd2a3cda87687e
2139 Merge: 4ad9716 6a4734a
2140 Author: Loic Dachary <loic@dachary.org>
2141 Date: Tue Nov 10 20:23:23 2015 +0100
2143 Merge pull request #5825 from tchaikov/wip-12848-hammer
2145 ReplicatedPG::hit_set_trim osd/ReplicatedPG.cc: 11006: FAILED assert(obc)
2147 Reviewed-by: Loic Dachary <ldachary@redhat.com>
2149 commit 2052187929e059a25f6a3baf67329f7ce0bf6d8a
2150 Author: Boris Ranto <branto@redhat.com>
2151 Date: Fri Oct 23 15:31:27 2015 +0200
2153 init-rbdmap: Rewrite to use logger + clean-up
2155 This patch rewrites the init-rbdmap init script so that it uses logger
2156 instead of the log_* functions. The patch also fixes various smaller
2158 * MAP_RV was undefined if mapping already existed
2159 * UMNT_RV and UMAP_RV were almost always empty (if they succeeded) ->
2161 * use of continue instead RET_OP in various places (RET_OP was not being
2162 checked after the switch to logger messages)
2163 * removed use of DESC (used only twice and only one occurrence actually
2166 Signed-off-by: Boris Ranto <branto@redhat.com>
2167 (cherry picked from commit c567341e98fffbe39177f951511a7698f88abf5f)
2169 commit f972f38445bc7a6efe122a5f9fc2ba7658625e26
2170 Author: Boris Ranto <branto@redhat.com>
2171 Date: Fri Oct 23 16:39:16 2015 +0200
2173 ceph.spec.in: We no longer need redhat-lsb-core
2175 Drop the redhat-lsb-core dependency as it is no longer necessary on
2178 The other two init scripts do not use redhat-lsb-core either. The
2179 init-ceph.in conditionally requires /lib/lsb/init-functions and does not
2180 use any of the functions defined in that file (at least not directly).
2181 The init-radosgw file includes /etc/rc.d/init.d/functions on non-debian
2182 platforms instead of /lib/lsb/init-functions file so it does not require
2183 redhat-lsb-core either.
2185 Signed-off-by: Boris Ranto <branto@redhat.com>
2186 (cherry picked from commit 15600572265bed397fbd80bdd2b7d83a0e9bd918)
2190 Merge conflict because there were other commits that introduced systemd
2191 & SUSE specific bits in the master which havent yet reached
2192 hammer. Since this commit only removed redhat-lsb-core from *el distros
2193 removing the specific lines
2195 commit 609f256a00f6ee18564ee055dd79f8293217940c
2196 Author: Jeff Weber <jweber@cofront.net>
2197 Date: Wed Aug 5 19:06:46 2015 -0400
2199 rgw: implement s3 encoding-type for get bucket
2201 This change introduces handling for the encoding-type request
2202 parameter on the get bucket operation. An object key may contain
2203 characters which are not supported in XML. Passing the value "url" for
2204 the encoding-type parameter will cause the key to be urlencoded in the
2208 Signed-off-by: Jeff Weber <jweber@cofront.net>
2209 (cherry picked from commit 180ca7b56ba91a3995c76dac698aa4ee31d9a2ce)
2211 commit 394fbfcc40d77ad6c8b3301c5ff2630c052107cd
2212 Author: Ruifeng Yang <yangruifeng.09209@h3c.com>
2213 Date: Fri Sep 25 10:18:11 2015 +0800
2215 Objecter: maybe access wild pointer(op) in _op_submit_with_budget.
2217 look at "after giving up session lock it can be freed at any time by response handler" in _op_submit,
2218 so the _op_submit_with_budget::op maybe is wild after call _op_submit.
2221 Signed-off-by: Ruifeng Yang <yangruifeng.09209@h3c.com>
2222 (cherry picked from commit 0635b1358354b19ae44105576f730381f3b5b963)
2224 commit 84068f8cea0452333f00f8b65230c00caa2e52ff
2225 Author: Ruifeng Yang <yangruifeng.09209@h3c.com>
2226 Date: Fri Sep 25 12:42:28 2015 +0800
2228 Objecter: repeated free op->ontimeout.
2230 repeated free op->ontimeout in SafeTimer::timer_thread::callback->complete
2233 Signed-off-by: Ruifeng Yang <yangruifeng.09209@h3c.com>
2234 (cherry picked from commit f1d8a8f577cee6d66f4dcffac667675f18145ebb)
2236 commit 3e657304dc41facd40b8cab7531180083d14d22a
2237 Author: Samuel Just <sjust@redhat.com>
2238 Date: Fri Aug 28 12:46:57 2015 -0700
2240 ReplicatedBackend::prepare_pull: set recover_info.size for clones
2243 Signed-off-by: Samuel Just <sjust@redhat.com>
2244 (cherry picked from commit 1d3e46bedb9a8c0d2200e39f62f4e2c3337619f3)
2246 commit 4ad97162026e1eb6e6e948ddf3eb39f711431e45
2247 Merge: 92b2153 f1271ea
2248 Author: Loic Dachary <loic@dachary.org>
2249 Date: Tue Nov 10 10:34:01 2015 +0100
2251 Merge pull request #6157 from dachary/wip-13040-hammer
2253 common/Thread:pthread_attr_destroy(thread_attr) when done with it
2255 Reviewed-by: Sage Weil <sage@redhat.com>
2257 commit 92b21532f64792da023188d9aa620d900c0245ca
2258 Merge: 1f02886 a17f2a9
2259 Author: Loic Dachary <loic@dachary.org>
2260 Date: Tue Nov 10 10:33:41 2015 +0100
2262 Merge pull request #6155 from dachary/wip-13171-hammer
2264 objecter: cancellation bugs
2266 Reviewed-by: Sage Weil <sage@redhat.com>
2268 commit 1f02886d7edc10ff93074afc3d2bf4cf77c709d5
2269 Merge: 77df385 7d4b303
2270 Author: Loic Dachary <loic@dachary.org>
2271 Date: Tue Nov 10 10:33:19 2015 +0100
2273 Merge pull request #6152 from dachary/wip-13233-hammer
2275 mon: include min_last_epoch_clean as part of PGMap::print_summary and PGMap::dump
2277 Reviewed-by: Sage Weil <sage@redhat.com>
2279 commit 77df385e23452f401f3c0c269e220446d288fb8e
2280 Merge: 469398f d8ca88d
2281 Author: Loic Dachary <loic@dachary.org>
2282 Date: Tue Nov 10 10:33:01 2015 +0100
2284 Merge pull request #6146 from dachary/wip-13337-hammer
2286 segfault in agent_work
2288 Reviewed-by: Sage Weil <sage@redhat.com>
2290 commit 469398f14b62af01068e65bcc5cc7c585a04e878
2291 Merge: d87d136 397042a
2292 Author: Loic Dachary <loic@dachary.org>
2293 Date: Tue Nov 10 10:32:53 2015 +0100
2295 Merge pull request #6145 from dachary/wip-13338-hammer
2297 filestore: fix peek_queue for OpSequencer
2299 Reviewed-by: Sage Weil <sage@redhat.com>
2301 commit d87d1364fb8e345a1f85b65c2d4c7c80596ba5f9
2302 Merge: 717eff9 d4e4d85
2303 Author: Loic Dachary <loic@dachary.org>
2304 Date: Tue Nov 10 10:32:42 2015 +0100
2306 Merge pull request #6144 from dachary/wip-13339-hammer
2308 mon: check for store writeablility before participating in election
2310 Reviewed-by: Sage Weil <sage@redhat.com>
2312 commit 717eff9a61818fde71ec5220998e9b5c14bbb3a9
2313 Merge: e133ddc e8cce08
2314 Author: Loic Dachary <loic@dachary.org>
2315 Date: Tue Nov 10 10:30:54 2015 +0100
2317 Merge pull request #5891 from Abhishekvrshny/wip-13035-hammer
2319 requeue_scrub when kick_object_context_blocked
2321 Reviewed-by: Sage Weil <sage@redhat.com>
2323 commit e133ddc622d3532b0182c208f0ebb9956daab44f
2324 Merge: c72306a 8ee93d5
2325 Author: Loic Dachary <loic@dachary.org>
2326 Date: Tue Nov 10 10:30:38 2015 +0100
2328 Merge pull request #5890 from Abhishekvrshny/wip-13037-hammer
2330 hit set clear repops fired in same epoch as map change -- segfault since they fall into the new interval even though the repops are cleared
2332 Reviewed-by: Sage Weil <sage@redhat.com>
2334 commit c72306a1dd3b193948d5df3371471797d4131cfe
2335 Merge: 5e8c9d9 2bd5d0b
2336 Author: Loic Dachary <loic@dachary.org>
2337 Date: Tue Nov 10 10:21:07 2015 +0100
2339 Merge pull request #6322 from sponce/WIP-13210-hammer
2341 tests : BACKPORT #13210 Fixed broken Makefiles after integration of ttng into rados
2343 Reviewed-by: Kefu Chai <kchai@redhat.com>
2345 commit b62cac6c345df53bbda1c42a5336cffc8f0f9652
2346 Author: Sangdi Xu <xu.sangdi@h3c.com>
2347 Date: Wed Sep 23 19:31:32 2015 +0800
2349 rgw: fix swift API returning incorrect account metadata
2353 Fix the bug that swift account stat command returns doubled object count and bytes used
2355 Signed-off-by: Sangdi Xu <xu.sangdi@h3c.com>
2356 (cherry picked from commit 66d19c78ba75b79190c81c95f94e7ef3084fda9e)
2358 commit 9643ee65a6e968e479ca33a102f5f575924a7ff0
2359 Author: Xinze Chi <xmdxcxz@gmail.com>
2360 Date: Sun Aug 2 18:36:40 2015 +0800
2362 bug fix: osd: do not cache unused buffer in attrs
2364 attrs only reference the origin bufferlist (decode from MOSDPGPush or
2365 ECSubReadReply message) whose size is much greater than attrs in recovery.
2366 If obc cache it (get_obc maybe cache the attr), this causes the whole origin
2367 bufferlist would not be free until obc is evicted from obc cache. So rebuild
2368 the bufferlist before cache it.
2371 Signed-off-by: Ning Yao <zay11022@gmail.com>
2372 Signed-off-by: Xinze Chi <xmdxcxz@gmail.com>
2373 (cherry picked from commit c5895d3fad9da0ab7f05f134c49e22795d5c61f3)
2375 commit 5e8c9d967e0ade393e36893965ca8ddfaa317b48
2376 Merge: db0366d dba8b5b
2377 Author: Loic Dachary <loic@dachary.org>
2378 Date: Mon Nov 9 08:38:31 2015 +0100
2380 Merge pull request #5810 from Abhishekvrshny/wip-12948-hammer
2382 Heavy memory shuffling in rados bench
2384 Reviewed-by: Kefu Chai <kchai@redhat.com>
2386 commit 1b06f03b541983461469f8f53ef2a271b08d2f78
2387 Author: Weijun Duan <duanweijun@h3c.com>
2388 Date: Thu Oct 29 21:46:02 2015 -0400
2390 rgw:swift use Civetweb ssl can not get right url
2394 Signed-off-by: Weijun Duan <duanweijun@h3c.com>
2395 (cherry picked from commit e0fd540bf441e2f8276cbd96c601a0539892efe2)
2397 commit db0366dd979cc0d401b1974c233e38dfe5b1b5d1
2398 Merge: 22dce75 a65c398
2399 Author: Loic Dachary <loic@dachary.org>
2400 Date: Thu Nov 5 13:12:11 2015 +0100
2402 Merge pull request #5530 from SUSE/wip-12587-hammer
2404 FileStore calls syncfs(2) even it is not supported
2406 Reviewed-by: Kefu Chai <kchai@redhat.com>
2408 commit 22dce75181b254548226fc09decafac14dcf84a6
2409 Merge: ed153c1 d8ac510
2410 Author: Loic Dachary <loic@dachary.org>
2411 Date: Wed Nov 4 19:38:53 2015 +0100
2413 Merge pull request #5889 from Abhishekvrshny/wip-13042-hammer
2415 ThreadPool add/remove work queue methods not thread safe
2417 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
2419 commit ed153c13226138eec71e472d1e3c8481db9b9009
2420 Merge: ca8802f 8610de8
2421 Author: Loic Dachary <loic@dachary.org>
2422 Date: Wed Nov 4 19:14:28 2015 +0100
2424 Merge pull request #6151 from dachary/wip-13245-hammer
2426 client nonce collision due to unshared pid namespaces
2428 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
2430 commit ca8802fb0cb8603267ecfc7320fee7a679617b8b
2431 Merge: 19450b9 6e29e90
2432 Author: Loic Dachary <loic@dachary.org>
2433 Date: Wed Nov 4 11:36:39 2015 +0100
2435 Merge pull request #6159 from dachary/wip-13341-hammer
2437 ceph upstart script rbdmap.conf incorrectly processes parameters
2439 Reviewed-by: Loic Dachary <ldachary@redhat.com>
2441 commit 19450b90e4c20556580d276bab7dafb23db1758f
2442 Merge: 18882c8 6849288
2443 Author: Loic Dachary <loic@dachary.org>
2444 Date: Wed Nov 4 11:34:07 2015 +0100
2446 Merge pull request #6154 from dachary/wip-13195-hammer
2448 should recalc the min_last_epoch_clean when decode PGMap
2450 Reviewed-by: Sage Weil <sage@redhat.com>
2452 commit 18882c81a71c4b517cb0eaa5a884637ea5cbe7a6
2453 Merge: de4f37b 3c1f7cb
2454 Author: Loic Dachary <loic@dachary.org>
2455 Date: Wed Nov 4 11:28:03 2015 +0100
2457 Merge pull request #6132 from SUSE/wip-13307-hammer
2459 dumpling incrementals do not work properly on hammer and newer
2461 Reviewed-by: Loic Dachary <ldachary@redhat.com>
2463 commit de4f37bb0abc932c813983792e170de1bb40241b
2464 Merge: 1c632a7 24268cf
2465 Author: Loic Dachary <loic@dachary.org>
2466 Date: Wed Nov 4 11:23:14 2015 +0100
2468 Merge pull request #6097 from dillaman/wip-13045-hammer
2470 librbd: diff_iterate needs to handle holes in parent images
2472 Reviewed-by: Loic Dachary <ldachary@redhat.com>
2474 commit 1c632a7ea7057522fb9df61994b5b3cf65416689
2475 Merge: bedb334 71a42f7
2476 Author: Loic Dachary <loic@dachary.org>
2477 Date: Wed Nov 4 11:22:25 2015 +0100
2479 Merge pull request #6060 from XinzeChi/wip-hammer-outdata-set
2481 bug fix: osd: avoid multi set osd_op.outdata in tier pool
2483 Reviewed-by: Loic Dachary <ldachary@redhat.com>
2485 commit bedb3344f6c802ea04a171dabd4fb596284af34a
2486 Merge: 0d3d819 cecd68d
2487 Author: Loic Dachary <loic@dachary.org>
2488 Date: Wed Nov 4 11:20:50 2015 +0100
2490 Merge pull request #5897 from tchaikov/wip-12940-hammer
2492 IO error on kvm/rbd with an erasure coded pool tier
2494 Reviewed-by: Loic Dachary <ldachary@redhat.com>
2496 commit 0d3d81951db61a97f6cb26d9f75e80a3400a2a37
2497 Merge: 381cfcc 7475a8f
2498 Author: Loic Dachary <loic@dachary.org>
2499 Date: Wed Nov 4 11:11:00 2015 +0100
2501 Merge pull request #5787 from SUSE/wip-12923-hammer
2503 logrotate reload error on Ubuntu 14.04
2505 Reviewed-by: Loic Dachary <ldachary@redhat.com>
2507 commit 381cfccf38bade28a001f3f24392c61896b3c711
2508 Merge: cad1cfb 86f88df
2509 Author: Loic Dachary <loic@dachary.org>
2510 Date: Wed Nov 4 11:10:26 2015 +0100
2512 Merge pull request #5766 from dachary/wip-12590-hammer
2514 ceph mds add_data_pool check for EC pool is wrong
2516 Reviewed-by: Greg Farnum <gfarnum@redhat.com>
2518 commit cad1cfb088b095e8333c4a7d79ccefc2b2c4d782
2519 Merge: 9764da5 e8d6d5a
2520 Author: Loic Dachary <loic@dachary.org>
2521 Date: Wed Nov 4 11:06:18 2015 +0100
2523 Merge pull request #5361 from ceph/wip-11786.hammer
2525 mon: MonitorDBStore: get_next_key() only if prefix matches
2527 Reviewed-by: Loic Dachary <ldachary@redhat.com>
2529 commit 5105d50d1c20d465229189bbb24cdfb5d16cf7bc
2530 Author: Loic Dachary <ldachary@redhat.com>
2531 Date: Tue Nov 3 00:21:51 2015 +0100
2533 tests: test/librados/test.cc must create profile
2535 Now that the create_one_ec_pool function removes the testprofile each
2536 time it is called, it must create the testprofile erasure code profile
2537 again for the test to use.
2539 http://tracker.ceph.com/issues/13664 Refs: #13664
2541 Signed-off-by: Loic Dachary <ldachary@redhat.com>
2542 (cherry picked from commit a60342942b5a42ee04d59af77a6b904ce62eefc4)
2544 commit 302375068a59b86c50bda73ecefd99831ab52ea4
2545 Author: Loic Dachary <ldachary@redhat.com>
2546 Date: Mon Nov 2 20:24:51 2015 +0100
2548 tests: destroy testprofile before creating one
2550 The testprofile erasure code profile is destroyed before creating a new
2551 one so that it does not fail when another testprofile erasure code
2552 profile already exists with different parameters.
2554 This must be done when creating erasure coded pools with the C++
2555 interface, in the same way it's done with the C interface.
2557 http://tracker.ceph.com/issues/13664 Fixes: #13664
2559 Signed-off-by: Loic Dachary <ldachary@redhat.com>
2560 (cherry picked from commit 47abab9a6f182aa0abe5047c04402850379bcd6d)
2562 commit 3a7423dbe0a6af831d4c47acc98e6ced5ed5fe0c
2563 Author: Loic Dachary <ldachary@redhat.com>
2564 Date: Mon Nov 2 20:23:52 2015 +0100
2566 tests: add destroy_ec_profile{,_pp} helpers
2568 Signed-off-by: Loic Dachary <ldachary@redhat.com>
2569 (cherry picked from commit ab46d79bc09fc711fa35302f49eecac81a98519b)
2571 commit a52f7cb372339dffbeed7dae8ce2680586760754
2572 Author: Sage Weil <sage@redhat.com>
2573 Date: Tue Oct 13 09:55:01 2015 -0400
2575 crush/mapper: ensure take bucket value is valid
2577 Ensure that the take argument is a valid bucket ID before indexing the
2580 Signed-off-by: Sage Weil <sage@redhat.com>
2581 (cherry picked from commit 93ec538e8a667699876b72459b8ad78966d89c61)
2583 commit 81d8aa14f3f2b7bf4bdd0b4e53e3a653a600ef38
2584 Author: Sage Weil <sage@redhat.com>
2585 Date: Tue Oct 27 20:55:26 2015 -0400
2587 crush/mapper: ensure bucket id is valid before indexing buckets array
2589 We were indexing the buckets array without verifying the index was within
2590 the [0,max_buckets) range. This could happen because a multistep rule
2591 does not have enough buckets and has CRUSH_ITEM_NONE
2592 for an intermediate result, which would feed in CRUSH_ITEM_NONE and
2596 Signed-off-by: Sage Weil <sage@redhat.com>
2597 (cherry picked from commit 976a24a326da8931e689ee22fce35feab5b67b76)
2599 commit 74203b806e8c268aa3edac8cfc7c4b393f7b1f27
2600 Author: xiexingguo <258156334@qq.com>
2601 Date: Mon Oct 26 18:38:01 2015 +0800
2603 FileStore: potential memory leak if _fgetattrs fails
2605 Memory leak happens if _fgetattrs encounters some error and simply returns.
2607 Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
2609 (cherry picked from commit ace7dd096b58a88e25ce16f011aed09269f2a2b4)
2611 commit db1cbe700e41ee7a2db5ef10173c429283bc5a2a
2612 Author: Robin H. Johnson <robin.johnson@dreamhost.com>
2613 Date: Fri Sep 4 01:07:48 2015 +0000
2615 Fix casing of Content-Type header
2617 It turns out, despite the HTTP spec declaring that header field names
2618 should be case-insensitive, some clients treat them wrongly, and
2619 consider "Content-type" to not match "Content-Type".
2621 CyberDuck was one of those clients, now fixed upstream in
2622 https://trac.cyberduck.io/ticket/8999
2624 To reduce future occurances of this bug, fix the casing of the
2625 Content-Type header, to strictly comply with the HTTP specification (be
2626 strict about what you send, and generous about what you receive).
2629 Backport: infernalis, hammer, firefly
2630 Signed-off-by: Robin H. Johnson <robin.johnson@dreamhost.com>
2631 (cherry picked from commit 1b9fbffdc24160251b96cec820d62fb2a12b6eab)
2633 commit 1448915e49bb6c8abc192053e6f2e8a380c4e92a
2634 Author: John Spray <john.spray@redhat.com>
2635 Date: Mon Jun 1 13:55:22 2015 +0100
2637 qa: update cephtool test for CephFS tier cases
2639 1. Creating a filesystem using a
2640 readonly tier on an EC pool (should be forbidden)
2641 2. Removing a tier from a replicated base pool (should
2644 Signed-off-by: John Spray <john.spray@redhat.com>
2645 (cherry picked from commit f9b11b0abe7c2b0dcdfc5802336be8a0426ccb3a)
2647 commit 14c36f0ccf103dea39b63e0e1b03082b99c42cb1
2648 Author: John Spray <john.spray@redhat.com>
2649 Date: Mon Jun 1 13:54:25 2015 +0100
2651 mon: forbid readonly cache tiers for CephFS
2653 For EC pools, we need a tier that is not just forwarding
2654 writes, in order to provide the ops required
2657 Signed-off-by: John Spray <john.spray@redhat.com>
2658 (cherry picked from commit 11e5faf3c1d4fa28a8b41e3a92f0f684a844e5f5)
2660 commit 31e59fbca0ba636fc57c208e612d7488a48a6287
2661 Author: John Spray <john.spray@redhat.com>
2662 Date: Thu May 7 14:23:37 2015 +0100
2664 mon: refine check_remove_tier checks
2667 Signed-off-by: John Spray <john.spray@redhat.com>
2668 (cherry picked from commit a50c8f1f2ad8845c7f77110868f9376f2d1ff883)
2670 commit 9f3aebee16e256888b149fa770df845787b06b6e
2671 Author: Sage Weil <sage@redhat.com>
2672 Date: Tue Sep 22 13:57:37 2015 -0400
2674 osd: fix requeue of replay requests during activating
2676 If the replay period expires while we are still in the activating
2677 state, we can simply insert our list of requests at the front of
2678 the waiting_for_active list.
2681 Signed-off-by: Sage Weil <sage@redhat.com>
2682 (cherry picked from commit d18cf51d9419819cdda3782b188b010969288911)
2684 commit ade0f1a4285348c20fbae8e2a366fc332a7b7fbe
2685 Author: xiexingguo <258156334@qq.com>
2686 Date: Tue Oct 13 14:04:20 2015 +0800
2688 OSD:shall reset primary and up_primary fields when beginning a new past_interval.
2690 Shall reset primary and up_primary fields when we start over a new past_interval in OSD::build_past_intervals_parallel().
2692 Signed-off-by: xie.xingguo@zte.com.cn
2693 (cherry picked from commit 65064ca05bc7f8b6ef424806d1fd14b87add62a4)
2695 commit 9085c820491f25a04ec02accc5098c1ab9b57311
2696 Author: yangruifeng <yangruifeng.09209@h3c.com>
2697 Date: Mon Oct 19 08:08:12 2015 -0400
2699 mon: should not set isvalid = true when cephx_verify_authorizer return false
2702 Signed-off-by: Ruifeng Yang <yangruifeng.09209@h3c.com>
2703 (cherry picked from commit c7f75b8f7c0a773148ec16141941efd00ee76626)
2705 commit 4875d05a3e265e80463a2a33a2c922affb0dad66
2706 Author: Ken Dreyer <kdreyer@redhat.com>
2707 Date: Mon Mar 16 10:02:35 2015 -0600
2709 ceph.spec.in: rm EOL Fedoras; add OBS RHEL5 instead
2711 Fedora 12 has been EOL for a long time. Remove the reference in the
2714 Since RHEL 5 support for Ceph is a work in progress, we won't remove
2715 this entire python_sitelib / python_sitearch conditional for now, since
2716 those are still needed on RHEL 5.
2718 Add the rhel_version macro to make the conditional compatible with
2721 Signed-off-by: Ken Dreyer <kdreyer@redhat.com>
2722 (cherry picked from commit 353a3258968f76deaea4542d2165e9ab32b88de8)
2724 commit dfba6acce75eea02251e98e2d57e6fbb344241fa
2725 Author: Travis Rhoden <trhoden@redhat.com>
2726 Date: Thu May 7 20:13:53 2015 -0400
2728 packaging: Add rgw placeholder dirs in proper packages
2730 Automatically create /var/lib/ceph/radosgw with the
2731 ceph-radosgw package, and /var/lib/ceph/bootstrap-rgw
2732 via the ceph package.
2734 Signed-off-by: Travis Rhoden <trhoden@redhat.com>
2735 (cherry picked from commit bf93128f664ac74383806530a1f974edde8df5b6)
2741 commit 2387c8e0caef00e7181241e16549457d58a12192
2742 Author: Nathan Cutler <ncutler@suse.cz>
2743 Date: Fri May 15 21:39:39 2015 +0200
2745 ceph.spec.in: consolidate centos/rhel macros
2747 0%{?rhel} || 0%{?centos} can be simplified to just 0%{?rhel}, since
2748 CentOS defines that macro for compatibility with RHEL.
2750 Signed-off-by: Nathan Cutler <ncutler@suse.cz>
2751 (cherry picked from commit 46404dd381e240634d224e77adc69f0ba4807045)
2755 https://github.com/ceph/ceph/pull/4068 has not been backported
2757 commit 7c58d312db2aa42088f419dbfa7e841d897fdad6
2758 Author: Nathan Cutler <ncutler@suse.com>
2759 Date: Thu Jun 25 16:55:10 2015 +0200
2761 packaging: make logrotate a hard dependency across all distros
2763 Given ceph's reputation for voluminous logs, it makes sense for
2764 logrotate to be a hard dependency for all distros.
2766 Signed-off-by: Nathan Cutler <ncutler@suse.com>
2767 (cherry picked from commit 0734cd1ec3ca3bd012686dc1bdda675909224a49)
2769 commit 61acc4e20ca58e11b9aba1efcdca78d1d7ddd2f7
2770 Author: Nathan Cutler <ncutler@suse.com>
2771 Date: Thu Jul 9 12:05:24 2015 +0200
2773 ceph.spec.in: add missing -%{release}
2775 We have it everywhere else and I can't think of any reason why
2776 it should be omitted here.
2778 Signed-off-by: Nathan Cutler <ncutler@suse.com>
2779 (cherry picked from commit d952d59ae95d5f66aa4b98fd7aa4f4ecb36a4848)
2781 commit 20590d367226ea01c9d4d95de15cfd1aa354d02b
2782 Author: Nathan Cutler <ncutler@suse.com>
2783 Date: Fri Jun 26 20:21:59 2015 +0200
2785 ceph.spec.in: remove duplicate BuildRequires: sharutils
2787 Signed-off-by: Nathan Cutler <ncutler@suse.com>
2788 (cherry picked from commit c418bd9615e6f23f3c14ae9e71003c66d31d4b78)
2790 commit 6b306eba829b1ec9822b24243ee7c021aab3d738
2791 Author: Nathan Cutler <ncutler@suse.com>
2792 Date: Mon Jun 15 12:58:00 2015 +0200
2794 ceph.spec.in: fix python-flask dependency for SUSE
2796 In SLE and openSUSE, the package is called python-Flask with an upper-case F.
2798 Signed-off-by: Nathan Cutler <ncutler@suse.com>
2799 (cherry picked from commit 39c28b9faf0fb069d91ac069a4d8df7bf03fb490)
2803 add missing SUSE-specific build deps
2805 commit 7a2cbcb0a4efa365b6d92ec58d7acf4a66e9d2b1
2806 Author: Nathan Cutler <ncutler@suse.com>
2807 Date: Thu Jun 25 18:55:39 2015 +0200
2809 ceph.spec.in: clarify two important comments
2811 First, the terms "common" and "specific" are vague. Second,
2812 "common" can easily be confused with the ceph-common subpackage.
2814 Fix this by rephrasing to "distro-unconditional dependencies" and
2815 "distro-conditional dependencies", respectively.
2817 Third, move the "distro-unconditional dependencies" header so it
2818 is above the part where these dependencies are actually defined.
2820 Signed-off-by: Nathan Cutler <ncutler@suse.com>
2821 (cherry picked from commit dd212fd1d367f70c61f65a97599959668b44aa53)
2823 commit dfe27c6ba05d49bb73854754b9d5c02cf1c7b1ef
2824 Author: Ken Dreyer <kdreyer@redhat.com>
2825 Date: Wed Jun 24 16:37:58 2015 -0600
2827 ceph.spec.in: rm reference to EOL Fedoras (< 20)
2829 Fedora 19 is EOL, so there's no need to refer to Fedoras less than 20.
2830 "%{fedora}" effectively implies "Fedora 20 and above".
2832 Reported-by: Ira Cooper <ira@samba.org>
2833 Signed-off-by: Ken Dreyer <kdreyer@redhat.com>
2834 (cherry picked from commit f3e4a91c458ca7b11706fe4608cb739499f5d3cc)
2838 added xmlstarlet dependency to ceph-test package
2840 commit 371b64ef4aa786db3fd29bc84ff98e63000a859b
2841 Author: Ken Dreyer <kdreyer@redhat.com>
2842 Date: Wed Jun 24 16:39:30 2015 -0600
2844 ceph.spec.in: package rbd-replay-prep on all Fedoras
2846 This reverts the change in commit
2847 85517d611b7bf4cb6cbffcd2c65303be0d038264. Since we BuildRequire:
2848 libbabeltrace-devel, autoconf will see that babeltrace is available
2849 during the build, and make will build/install the rbd-replay-prep
2852 This change also simplifies Fedora selection logic, because Fedora 19 is
2853 EOL, so "%{fedora}" implies "Fedora 20 and above".
2855 Reported-by: Ira Cooper <ira@samba.org>
2856 Signed-off-by: Ken Dreyer <kdreyer@redhat.com>
2857 (cherry picked from commit 2db9480fa2646de776f0076d82747e202e574ddd)
2859 commit 82425b33f712047442d9d29c5a4349818b44b40a
2860 Author: Boris Ranto <branto@redhat.com>
2861 Date: Tue Jun 16 23:07:04 2015 +0200
2863 ceph.spec.in: rbd-replay-prep is not being built on f21+
2865 Signed-off-by: Boris Ranto <branto@redhat.com>
2866 (cherry picked from commit 85517d611b7bf4cb6cbffcd2c65303be0d038264)
2868 commit 53e2f702681dc08ddd0414b116e51de6488f42ca
2869 Author: Owen Synge <osynge@suse.com>
2870 Date: Mon Jun 8 17:48:55 2015 +0200
2872 ceph.spec.in:BuildRequires sharutils
2874 The uudecode binary is used to build Java-related components, and
2875 uudecode is provided by the sharutils package on all supported
2876 RPM platforms. When building with "--without=cephfs_java",
2877 sharutils is not needed.
2879 Thanks to Nathan Cutler <ncutler@suse.cz> for going into the
2882 On OBS without this patch we get the error message:
2884 [ 170s] -----------------------------------------------------------------
2885 [ 170s] ----- building ceph.spec (user abuild)
2886 [ 170s] -----------------------------------------------------------------
2887 [ 170s] -----------------------------------------------------------------
2888 [ 170s] + exec rpmbuild -ba --define '_srcdefattr (-,root,root)' --nosignature --define 'disturl obs://build.opensuse.org/home:osynge:ceph:wip:wip_obs_fedora/Fedora_20/c0bbbc1e62228ca956ac3d367edc4fba-master' /home/abuild/rpmbuild/SOURCES/ceph.spec
2889 [ 170s] error: Failed build dependencies:
2890 [ 170s] sharutils is needed by ceph-1:2+git.1435043747.c1bd02c-1.1.x86_64
2892 With this patch we can build fedora 22 and fedora 20 rpms fine.
2894 Signed-off-by: Owen Synge <osynge@suse.com>
2895 (cherry picked from commit 43c1784640538b813f0575c50f721ac911a30b3e)
2899 _with_systemd -> trivial resolution
2901 commit a84376c68499845de01f9e3b194b16a698125b6a
2902 Author: Owen Synge <osynge@suse.com>
2903 Date: Wed Jun 10 12:11:50 2015 +0200
2905 ceph.spec.in: fix:Add missing directories breaking build
2907 SUSE builds on OBS are failing with the missing dir entries:
2912 On suse these correspond to:
2914 %dir %{_datarootdir}/ceph/
2915 %dir %{_libexecdir}/ceph/
2917 Signed-off-by: Owen Synge <osynge@suse.com>
2918 (cherry picked from commit f1ff3e4dcb5dd7eddd2c346a05cfa78d860a5ce6)
2920 commit 171fee1b82d2675e364da7f96dfb9dd286d9b6e6
2921 Author: Kefu Chai <kchai@redhat.com>
2922 Date: Tue Jul 21 00:33:19 2015 +0800
2924 mon: track osd_epoch of all MonSessions
2926 previously, we only track the osd_epoch for OSD peers. but other
2927 MonClients who receives osdmaps can also benefit from osd_epoch.
2929 Signed-off-by: Kefu Chai <kchai@redhat.com>
2930 (cherry picked from commit de43a02e06650a552f048dc8acd17f255126fed9)
2933 src/mon/OSDMonitor.cc:
2934 do not assume that the MonSession is always available when
2935 OSDMonitor::send_incremental() is called. as when the
2936 connection to the mon client is dropped, the Monitor is
2937 aware of this, but not the OSDMonitor.
2939 commit cc7da674965c3f758469e18788dc0e106e54d5bc
2940 Author: Kefu Chai <kchai@redhat.com>
2941 Date: Tue Jul 21 00:24:52 2015 +0800
2943 mon: track osd_epoch in MonSession
2945 * remove osd_epoch<osd, epoch> from OSDMonitor
2946 * add osd_epoch to MonSession to track the latest osdmap epoch
2947 OSDMonitor sends to a mon client
2948 * do not remove osd_epoch entries if an OSD is down, or
2952 Signed-off-by: Kefu Chai <kchai@redhat.com>
2953 (cherry picked from commit c05753eacc26e90b2e3b56e641a71bffd5b39bd0)
2955 commit 3f33ce6132d5ce8b3454f30fbe14471ece39951d
2956 Author: Lu Shi <shi.lu@h3c.com>
2957 Date: Thu Sep 24 16:02:41 2015 +0800
2959 librbd:reads larger than cache size hang.
2963 Signed-off-by: Lu Shi <shi.lu@h3c.com>
2964 (cherry picked from commit 9c8200bb5d1ac9359803a182df03298b565b8479)
2966 commit 5e5b512189b809bef503c34b0a19e30ed81cfbfa
2967 Author: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2968 Date: Wed Aug 19 20:32:39 2015 +0200
2970 rgw: url_decode values from X-Object-Manifest during GET on Swift DLO.
2973 Signed-off-by: Radoslaw Zarzynski <rzarzynski@mirantis.com>
2974 (cherry picked from commit 69cf089596a0e5d4da2826b5c276306316a00690)
2976 commit e693755a55e4da681a1947b1e1eb3bb537436cc0
2977 Author: Yehuda Sadeh <yehuda@redhat.com>
2978 Date: Tue Sep 22 14:53:42 2015 -0700
2980 cls_rgw: fix bucket listing when dealing with invisible entries
2984 The bucket listing logic was broken when dealing with invisible entries.
2985 Need to read keys until we can fetch all the requested entries, or
2986 determine that there are no more entries. Anything else requires
2987 a change to the protocol.
2989 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
2990 (cherry picked from commit ef9730515e001d77637a77eb89950f70fd5e7f5a)
2992 commit 99b4d1df3f6867e5ff0dd74691203737f6de6974
2993 Author: Yehuda Sadeh <yehuda@redhat.com>
2994 Date: Tue Sep 22 15:49:27 2015 -0700
2996 rgw: orphan tool shouldn't clean up head objects
3000 Head objects are mutable, so removing them can race with object removal
3001 and a later recreation, so we might end up cleaning them up when we don't
3004 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
3005 (cherry picked from commit 030f697d196df9963ecf4f05f727da4798c583dc)
3007 commit 6d89f4ba85063debe96734e9fdfcba247e770206
3008 Author: Yehuda Sadeh <yehuda@redhat.com>
3009 Date: Mon Oct 12 11:55:40 2015 -0700
3011 rgw: location constraints should return api name
3014 The S3 location constraints request should return the api name,
3015 not the region name.
3017 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
3018 (cherry picked from commit 03078bab6ee8c617025902dbe52ec19e64653c33)
3020 commit 2bd5d0b835b546c80b5af503d2b9538835c768fd
3021 Author: Sebastien Ponce <sebastien.ponce@cern.ch>
3022 Date: Tue Oct 20 09:22:16 2015 +0200
3024 radosstriper : Fixed broken Makefiles after integration of lttng
3026 Fixes: #13210, backport of #6322 into hammer
3028 Signed-off-by: Sebastien Ponce <sebastien.ponce@cern.ch>
3030 commit 7ffd072a8cc7e114d4d674d30333fea4db04dcdd
3031 Author: Sage Weil <sage@redhat.com>
3032 Date: Thu Oct 1 14:50:34 2015 -0400
3034 osdc/Objecter: distinguish between multiple notify completions
3036 We may send a notify to the cluster multiple times due to OSDMap
3037 changes. In some cases, earlier notify attempts may complete with
3038 an error, while later attempts succeed. We need to only pay
3039 attention to the most-recently send notify's completion.
3041 Do this by making note of the notify_id in the initial ACK (only
3042 present when talking to newer OSDs). When we get a notify
3043 completion, match it against our expected notify_id (if we have
3044 one) or else discard it.
3046 This is important because in some cases an early notify completion
3047 may be an error while a later one succeeds.
3049 Note that if we are talking to an old cluster we will simply not record a
3050 notify_id and our behavior will be the same as before (we will trust any
3051 notify completion we get).
3054 Signed-off-by: Sage Weil <sage@redhat.com>
3055 (cherry picked from commit e86d033854c76f344c678e92016c4e5c5e0385e2)
3058 src/osdc/Objecter.cc
3059 In Objecter::handle_watch_notify, a conflict was there due to a modified comment by commit 47277c51db7bb2725ea117e4e8834869ae93e006, which was not backported
3061 commit 647c91098e8517912c73e3f7d27954e6f340472c
3062 Author: Sage Weil <sage@redhat.com>
3063 Date: Thu Oct 1 14:50:00 2015 -0400
3065 osd: reply to notify request with our unique notify_id
3067 The OSD assigns a unique ID to each notify it queues for
3068 processing. Include this in the reply to the notifier so that
3069 they can match it up with the eventual completions they receive.
3071 This is necessary to distinguish between multiple completions
3072 they may receive if there is PG peering and the notify is resent.
3073 In particular, an earlier notify may return an error when a later
3076 This is forwards and backwards compatible: new clients will make use of
3077 this reply payload but older clients ignore it.
3079 Signed-off-by: Sage Weil <sage@redhat.com>
3080 (cherry picked from commit 049ea702b9bf4a8a62ae2770d9ba55c0730b3eef)
3082 commit 0ad9521944e0de5c135a066b0cda07f8976bdae1
3083 Author: David Zafman <dzafman@redhat.com>
3084 Date: Mon Oct 12 13:24:31 2015 -0700
3086 osd: Correct the object_info_t::decode() version
3091 Backport: Hammer, Firefly
3093 Signed-off-by: David Zafman <dzafman@redhat.com>
3094 (cherry picked from commit 2b7ddde685213f3908f2dd9545b35f1d97f35a79)
3096 commit af734e63f28ac5af414b6507c945ffd9e736d5cf
3097 Author: Jason Dillaman <dillaman@redhat.com>
3098 Date: Tue Oct 6 12:31:59 2015 -0400
3100 librbd: invalidate object map on error even w/o holding lock
3102 If there is a deep bug that prevents the object map from owning
3103 the lock before objects are updated, flag the object map as invalid
3108 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
3109 (cherry picked from commit e675400df7f8ba796d60e9bac4234857f0cb1392)
3112 src/librbd/ObjectMap.h/cc: hammer branch doesn't support rebuild
3114 commit 634d7f62486847846e3862091b8cb587d095cb52
3115 Author: Jason Dillaman <dillaman@redhat.com>
3116 Date: Thu Oct 8 13:28:04 2015 -0400
3118 tests: removed obsolete rbd_replay test cases
3120 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
3121 (cherry picked from commit e0d8cb1283e4bdbfc770684b173e477c22558524)
3123 commit 5c63d3c614189381cc169c64a3138c7abcb368c6
3124 Author: Jason Dillaman <dillaman@redhat.com>
3125 Date: Thu Oct 8 13:26:42 2015 -0400
3127 ceph-dencoder: new rbd_replay trace file types
3129 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
3130 (cherry picked from commit c2a83d0b518497dc8bcbbdea77b88febe52a3034)
3132 commit a145a1cd8e245505abae66f0497be0df0078cecb
3133 Author: Jason Dillaman <dillaman@redhat.com>
3134 Date: Thu Oct 8 13:21:29 2015 -0400
3136 rbd-replay: added version control to trace output file
3138 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
3139 (cherry picked from commit 3ecdae8388d69123b937a40ce614a0b795a757f1)
3141 commit a30ed76c34ecca6278112908871126e1730a22bc
3142 Author: Jason Dillaman <dillaman@redhat.com>
3143 Date: Wed Oct 7 15:39:18 2015 -0400
3145 rbd-replay-prep: added --verbose command line option
3147 This new command-line will dump all the processed events as they
3148 are extracted from the trace file.
3150 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
3151 (cherry picked from commit 646e50a771c3089121ec2f4369d1a8188001a3eb)
3153 commit adf31c88d3ed95d0c29c649e2642ee3ae758743b
3154 Author: Jason Dillaman <dillaman@redhat.com>
3155 Date: Wed Oct 7 15:22:25 2015 -0400
3157 rbd-replay-prep: stream events to the prep file
3159 Avoid building the entire prep event history in memory before
3160 attempting to write the prepped trace file.
3162 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
3163 (cherry picked from commit 98f513a113f4e8aff17c83991d2e6f0f3738a9c9)
3165 commit 8c9e67ad0b21629565a09f6b9b92723b98594f84
3166 Author: Jason Dillaman <dillaman@redhat.com>
3167 Date: Wed Oct 7 14:56:22 2015 -0400
3169 rbd-replay-prep: simplify IO dependency calculation
3171 Only track read-after-write and write-after-write IO dependencies
3172 via the associated write completions. All IO events after a write
3173 completion are considered to be dependent and can be pruned down
3174 to at most the number of concurrent IOs. This reduces the prep
3175 time from a simple 'rbd bench-write' from over 4 hrs down to seconds.
3177 Fixes: #13378, #13384
3180 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
3181 (cherry picked from commit 65fb1b86cbab65023f6207798c9e189bce55dcf6)
3183 commit e7406fcba86646a4c3fbac66ee54b52ce3d5a33d
3184 Author: Jason Dillaman <dillaman@redhat.com>
3185 Date: Tue Oct 6 13:16:51 2015 -0400
3187 rbd-replay: improve error messages
3191 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
3192 (cherry picked from commit 4f436e10802d7b3c3949b592fec238c189a02c50)
3194 commit 9f81d28107bd82b7118d1521ade982ab179c5a7a
3195 Author: Jason Dillaman <dillaman@redhat.com>
3196 Date: Tue Oct 6 12:48:50 2015 -0400
3198 rbd-replay-prep: support new read/write APIs
3200 Added support for read2, write2, aio_read2, and aio_write2 APIs.
3204 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
3205 (cherry picked from commit b5b4a9dd6bfe10844327a457849a0dbf7ef597ea)
3207 commit a856bee2aa7e22f2bedf54b684c4d55452766b01
3208 Author: Jason Dillaman <dillaman@redhat.com>
3209 Date: Mon Oct 5 17:30:54 2015 -0400
3211 rbd-replay-prep: avoid using assert statements on user inputs
3215 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
3216 (cherry picked from commit a1e99f06a78d7cf2ecc0c8e9fbc3331e4e4f18f7)
3218 commit 6a4734a7afecb359b8bc9c114332cd7321097c38
3219 Author: Samuel Just <sjust@redhat.com>
3220 Date: Mon Sep 21 11:16:49 2015 -0700
3222 ReplicatedPG::hit_set_setup: fix hit_set_remove_all call
3224 We only want to do it if the pool config changed AND we are
3227 fd38902dd4693b9b72f7532833e78f5df2c9efa5 partially fixed a related
3228 bug. This should be backported along with it.
3231 Signed-off-by: Samuel Just <sjust@redhat.com>
3232 (cherry picked from commit 36e4a80c39f7daab2a35efc04650ae8b7b11c4e0)
3234 commit ab3877dd566f3951e76abaa6b7deeb5835a512b1
3235 Author: Samuel Just <sjust@redhat.com>
3236 Date: Tue Sep 22 11:13:44 2015 -0700
3238 osd/: assert in HitSet constructor if type is TYPE_NONE as well
3240 Otherwise impl is unpopulated and we risk segfaults in several
3241 methods. It also seems like it would always indicate a bug.
3243 Signed-off-by: Samuel Just <sjust@redhat.com>
3244 (cherry picked from commit 8e5a801b43babf751b619ed9606d7be7e5b12837)
3246 commit 4fe07b7f69321535a985958fc761a54cb71136f6
3247 Author: Zhiqiang Wang <zhiqiang.wang@intel.com>
3248 Date: Thu Jul 23 09:26:44 2015 +0800
3250 osd: only remove all the hit set objects on primary osd
3252 Only the primary PG is allowed to remove all the hit set objects. And
3253 the PG should be in the active or peered states.
3255 Signed-off-by: Zhiqiang Wang <zhiqiang.wang@intel.com>
3256 (cherry picked from commit fd38902dd4693b9b72f7532833e78f5df2c9efa5)
3258 commit 313a09b58b6aa6fe8999f4d5da07a04f00c64cd7
3259 Author: Zhiqiang Wang <zhiqiang.wang@intel.com>
3260 Date: Thu Jun 18 09:05:28 2015 +0800
3262 osd: implement hit_set_remove_all
3264 When hit set is not configured on startup or on a change, remove all
3267 Signed-off-by: Zhiqiang Wang <zhiqiang.wang@intel.com>
3268 (cherry picked from commit be28319bf3dc54b4b6f400d83405a29facfe3bd4)
3270 src/osd/ReplicatedPG.cc
3271 pass the new argument of `p->using_gmt` to get_hit_set_archive_object()
3273 commit b632254252f56d3b2c0c32803593ba765b45b222
3274 Author: Kefu Chai <kchai@redhat.com>
3275 Date: Tue Sep 8 16:52:32 2015 +0800
3277 mon: disable gmt_hitset if not supported
3279 the gmt_hitset is enabled by default in the ctor of pg_pool_t, this
3280 is intentional. because we want to remove this setting and make
3281 gmt_hitset=true as a default in future. but this forces us to
3282 disable it explicitly when preparing a new pool if any OSD does
3283 not support gmt hitset.
3286 Signed-off-by: Kefu Chai <kchai@redhat.com>
3287 (cherry picked from commit e48cec3dc93b3988dcd8924933deb1b3a43e1d0f)
3290 src/mon/OSDMonitor.cc
3291 remove unrelated hunk
3293 commit 27fc9a3b2bd34bec8d5bb3954b873ce6344771f4
3294 Author: Kefu Chai <kchai@redhat.com>
3295 Date: Fri Aug 28 00:32:38 2015 +0800
3297 osd: do not let OSD_HITSET_GMT reuse the feature bit
3299 * to ease the backport to hammer
3301 Signed-off-by: Kefu Chai <kchai@redhat.com>
3302 (cherry picked from commit 5a4f6a866bb675195ad83c38fd23238d078c78ed)
3305 src/include/ceph_features.h
3306 minor changes to resolve conflicts
3308 commit 7fcc61d26028eb417a0c224e8c9cd3cc95bc13d9
3309 Author: David Zafman <dzafman@redhat.com>
3310 Date: Thu Aug 27 11:24:25 2015 -0700
3312 osd: Decode use_gmt_hitset with a unique version
3314 Signed-off-by: David Zafman <dzafman@redhat.com>
3315 (cherry picked from commit 2bc5a48f4c5d3667213be3a7b5a0e0f5ef9daf4f)
3318 src/osd/osd_types.cc
3319 minor changes to resolve conflicts
3321 commit 26c7e968523ac09f608c4ed7c50681fbb741ca1f
3322 Author: Kefu Chai <kchai@redhat.com>
3323 Date: Tue Aug 18 16:53:50 2015 +0800
3325 mon: print use_gmt_hitset in "ceph osd pool get"
3327 Signed-off-by: Kefu Chai <kchai@redhat.com>
3328 (cherry picked from commit cc2bcf760f2d2f20fc4a2fa78ba52475a23e55bf)
3331 src/mon/OSDMonitor.cc
3332 use the old way to dump pool info
3334 commit e8e00dab1b064e77a997504f385b5d60ee8aca25
3335 Author: Kefu Chai <kchai@redhat.com>
3336 Date: Tue Aug 18 00:04:23 2015 +0800
3338 mon: add "ceph osd pool set $pool use_gmt_hitset true" cmd
3340 allow "ceph osd pool set $pool use_gmt_hitset <true|1>" as long as
3341 the cluster supports gmt hitset.
3344 Signed-off-by: Kefu Chai <kchai@redhat.com>
3345 (cherry picked from commit 03a1a3cf023a9aeb2fa26820e49e5efe3f3b3789)
3347 commit 040e390d29fc68dcefe48c59cccacf6258c6f690
3348 Author: Kefu Chai <kchai@redhat.com>
3349 Date: Fri Jun 5 21:06:48 2015 +0800
3351 osd: use GMT time for the object name of hitsets
3353 * bump the encoding version of pg_hit_set_info_t to 2, so we can
3354 tell if the corresponding hit_set is named using localtime or
3356 * bump the encoding version of pg_pool_t to 20, so we can know
3357 if a pool is using GMT to name the hit_set archive or not. and
3358 we can tell if current cluster allows OSDs not support GMT
3360 * add an option named `osd_pool_use_gmt_hitset`. if enabled,
3361 the cluster will try to use GMT mode when creating a new pool
3362 if all the the up OSDs support GMT mode. if any of the
3363 pools in the cluster is using GMT mode, then only OSDs
3364 supporting GMT mode are allowed to join the cluster.
3367 Signed-off-by: Kefu Chai <kchai@redhat.com>
3368 (cherry picked from commit 42f8c5daad16aa849a0b99871d50161673c0c370)
3371 src/include/ceph_features.h
3372 src/osd/ReplicatedPG.cc
3373 src/osd/osd_types.cc
3375 fill pg_pool_t with default settings in master branch.
3377 commit 31b7864b4161221f6027c6cbe1e42ed598d42b6a
3378 Author: Nathan Cutler <ncutler@suse.com>
3379 Date: Sun Sep 27 21:55:00 2015 +0200
3381 CephxServiceHandler.cc: fix get_auth conditional
3383 The KeyServer class has a public method get_auth() that returns a boolean
3384 value. This value is being checked here - fix the conditional so it triggers
3385 when get_auth() returns false.
3387 http://tracker.ceph.com/issues/9756 References: #9756
3389 Signed-off-by: Nathan Cutler <ncutler@suse.com>
3390 (cherry picked from commit 7250db62cb7ac1e9c0b1f5956d60b9d874496702)
3392 commit 0742177c767613d41bda0d260eebdd0087e3b392 (refs/remotes/me/wip-randomize-scrub-hammer)
3393 Author: Kefu Chai <kchai@redhat.com>
3394 Date: Sat May 16 00:07:27 2015 +0800
3396 osd: use another name for randomize scrub option
3398 s/osd_scrub_interval_limit/osd_scrub_interval_randomize_ratio/
3401 Signed-off-by: Kefu Chai <kchai@redhat.com>
3403 commit fad33861a87f2feb944ad5454775c95d09364b56
3404 Author: Kefu Chai <kchai@redhat.com>
3405 Date: Mon Mar 9 16:42:34 2015 +0800
3407 osd: randomize scrub times to avoid scrub wave
3409 - to avoid the scrub wave when the osd_scrub_max_interval reaches in a
3410 high-load OSD, the scrub time is randomized.
3411 - extract scrub_load_below_threshold() out of scrub_should_schedule()
3412 - schedule an automatic scrub job at a time which is uniformly distributed
3413 over [now+osd_scrub_min_interval,
3414 now+osd_scrub_min_interval*(1+osd_scrub_time_limit]. before
3415 this change this sort of scrubs will be performed once the hard interval
3416 is end or system load is below the threshold, but with this change, the
3417 jobs will be performed as long as the load is low or the interval of
3418 the scheduled scrubs is longer than conf.osd_scrub_max_interval. all
3419 automatic jobs should be performed in the configured time period, otherwise
3421 - the requested scrub job will be scheduled right away, before this change
3422 it is queued with the timestamp of `now` and postponed after
3423 osd_scrub_min_interval.
3426 Signed-off-by: Kefu Chai <kchai@redhat.com>
3427 (cherry picked from commit 5e44040e8528bff06cc0a5a3f3293ab146e0e4e1)
3432 commit 9c130336a1a1dcf2338d202914aeacbc194af0b0
3433 Author: Kefu Chai <kchai@redhat.com>
3434 Date: Mon Mar 9 16:47:30 2015 +0800
3436 osd: use __func__ in log messages
3438 Signed-off-by: Kefu Chai <kchai@redhat.com>
3439 (cherry picked from commit 0f7f35670f03f3f58329c00d323963710e7e495d)
3441 commit 4a3dedff8693953944af3eeb40994f38befebf48
3442 Author: Kefu Chai <kchai@redhat.com>
3443 Date: Wed Apr 15 12:04:05 2015 +0800
3445 osd: simplify OSD::scrub_load_below_threshold() a little bit
3447 avoid unnecessary comparison
3449 Signed-off-by: Kefu Chai <kchai@redhat.com>
3450 (cherry picked from commit 2ab0e606dfd2847e1918be9f21f572d91befbf7b)
3452 commit 24268cf4309267a0a1e32e28379cf7eafac00bca
3453 Author: Jason Dillaman <dillaman@redhat.com>
3454 Date: Fri May 1 14:21:35 2015 -0400
3456 librbd: invoking RBD::open twice will leak memory
3458 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
3459 (cherry picked from commit a5e88fcf11ebb16d4a1405b21cf6f895b37f185e)
3461 commit 8610de81c8eeae763354a4133e160ccb3faa8dad
3462 Author: Josh Durgin <jdurgin@redhat.com>
3463 Date: Mon Sep 21 15:46:42 2015 -0700
3465 use simplifed messenger constructor for clients
3467 This is all mechanical except the calculation of the nonce, which is
3468 now always randomized for clients.
3471 Signed-off-by: Josh Durgin <jdurgin@redhat.com>
3472 (cherry picked from commit a3a8c85b79afef67681c32c57b591c0e0a87a349)
3474 commit 9d059c3f1b3aa0b1830b6f84694c850d36e8a77a
3475 Author: Josh Durgin <jdurgin@redhat.com>
3476 Date: Mon Sep 21 15:20:53 2015 -0700
3478 msg: add simplified messenger constructor
3480 Several callers create messengers using exactly the same parameters:
3482 - reading the ms type from cct that is also passed in
3483 - a default entity_name_t::CLIENT
3484 - the default features
3486 Additionally, the nonce should be randomized and not depend on
3487 e.g. pid, as it does in several callers now. Clients running in
3488 containers can easily have pid collisions, leading to hangs, so
3489 randomize the nonce in this simplified constructor rather than
3490 duplicating that logic in every caller.
3492 Daemons have meaningful entity_name_ts, and monitors currently depend
3493 on using 0 as a nonce, so make this simple constructer
3497 Signed-off-by: Josh Durgin <jdurgin@redhat.com>
3498 (cherry picked from commit e3785b0b5fbff870adbd5f121021b671912386cf)
3501 src/msg/Messenger.cc: Messenger::create only has 5 arguments in
3502 hammer. Since the last one is zero, it does not change the
3505 commit 6e29e904ce8e7b3481fd3b318b070d326cd4a71a
3506 Author: Sage Weil <sage@redhat.com>
3507 Date: Wed Sep 30 08:29:05 2015 -0400
3509 init-rbdmap: fix CMDPARAMS
3512 Reported-by: Wyllys Ingersoll <wyllys.ingersoll@keepertech.com>
3513 Signed-off-by: Sage Weil <sage@redhat.com>
3514 (cherry picked from commit 02113ac449cd7631f1c9a3840c94bbf253c052bd)
3516 commit 7161a2c927a6ded0d6ffe7d7621b7abd7f18acc3
3517 Author: Samuel Just <sjust@redhat.com>
3518 Date: Mon Sep 21 12:00:49 2015 -0700
3520 PG: ignore info from down osd
3523 Backport: firefly, hammer
3524 Signed-off-by: Samuel Just <sjust@redhat.com>
3525 (cherry picked from commit 930d8eb1fb0daacd692d2e80b5ff473448bd4e8d)
3527 commit 9753551f9f4ea9a1638dd469ef536d895ca9bc3b
3528 Author: Samuel Just <sjust@redhat.com>
3529 Date: Mon Sep 21 12:01:03 2015 -0700
3531 OSDMap.h: add has_been_up_since
3533 Signed-off-by: Samuel Just <sjust@redhat.com>
3534 (cherry picked from commit 4b00350870e3a43aee5fff7982b3785fe1876634)
3536 commit f1271eae0decf4d1fd9cfd83e5e54d3be6abfb17
3537 Author: zqkkqz <zheng.qiankun@h3c.com>
3538 Date: Fri Aug 7 10:49:45 2015 +0800
3540 Common/Thread: pthread_attr_destroy(thread_attr) when done with it
3541 When a thread attributes object is no longer required, it should be destroyed using the
3542 pthread_attr_destroy() function. Destroying a thread attributes object has no effect on threads that were created using that object.
3545 Signed-off-by: zqkqkz zheng.qiankun@h3c.com
3546 (cherry picked from commit 9471bb838a420ef5f439191d87e5388fecccb9e6)
3548 commit 9581f9f4446ed5838bb3de6e1d7cf5faca85df90
3549 Author: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
3550 Date: Fri Jul 17 12:43:52 2015 +0200
3552 Thread.cc: remove malloc/free pair
3554 There's no need for mallocing pthread_attr_t in Thread::try_create(),
3555 it can be located on stack as it is freed in same function. This reduces
3556 pressure put on memory manager.
3558 Signed-off-by: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
3559 (cherry picked from commit 614ac0f4e375f3dd008c8937d3f89f6005d6e0a0)
3561 commit a17f2a9a6600803a8b7f6dd306f7bafb6acc1af5
3562 Author: Jianpeng Ma <jianpeng.ma@intel.com>
3563 Date: Tue Sep 1 21:35:19 2015 +0800
3565 osdc/Objecter: optimize Objecter::tick.
3567 Set bool value is better than insert same value into set<>.
3569 Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
3570 (cherry picked from commit ee204044d9e69ea44b533c05cec154974039264c)
3572 commit f80d237decac775d82f1e276a1ea600a9f352ffe
3573 Author: Jianpeng Ma <jianpeng.ma@intel.com>
3574 Date: Tue Sep 1 19:32:32 2015 +0800
3576 osdc/Objecter: remove the unuseful code.
3578 Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
3579 (cherry picked from commit 80f10e3e59dadda2dca4eb62c68af972b701b316)
3581 commit 8f76f4c467dc69cd2af75d9cbb134c26f2d04a84
3582 Author: Jianpeng Ma <jianpeng.ma@intel.com>
3583 Date: Tue Sep 1 15:39:29 2015 +0800
3585 osdc/Objecter: Don't forget call _op_cancel_map_check when cancel linger op.
3587 Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
3588 (cherry picked from commit 7cc963b1c6ab37bf33638dc6eca7848d93f7908f)
3590 commit ba36caba5226a44c41fdd2b19074916b38af5fdf
3591 Author: Jianpeng Ma <jianpeng.ma@intel.com>
3592 Date: Tue Sep 1 15:12:02 2015 +0800
3594 osdc/Objecter: In _cancel_linger_op, it should make num_unacked/num_committed decrease.
3596 Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
3597 (cherry picked from commit 36b62710ddef0ffaee25837a92ca1ac9b353ff05)
3599 commit 6ff7b02dc23b42a99c7136d1a55ead05c12e2b97
3600 Author: Jianpeng Ma <jianpeng.ma@intel.com>
3601 Date: Thu Aug 20 17:00:23 2015 +0800
3603 osdc/Objeter: When cancel op, decrease num_unacked/num_uncommitted.
3605 Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
3606 (cherry picked from commit 064e8585a04edb3d87b38db6bed03e965cfcb359)
3608 commit d815e0aacb4a516f4bbbb59a2b2073b38759d9e8
3609 Author: Jianpeng Ma <jianpeng.ma@intel.com>
3610 Date: Thu Aug 20 15:38:58 2015 +0800
3612 osdc/Objecter: For func op_cancel_writes it can directly call op_cancel.
3614 Becasue we get write-lock of rwlock, so it is safe to call op_cancel
3615 rather than _op_canchel(homeless_session for this case don't met).
3617 Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
3618 (cherry picked from commit e4ce619fe17a7a9dfc18e6af0b84928aa2d88c00)
3621 src/osdc/Objecter.cc: the last if (found) was incorrectly
3622 implemented as if (to_cancel.size()) after hammer. The
3623 conflict is because the patch tries to revert an incorrect
3624 if (to_cancel.size()) which is not in hammer.
3626 commit 53316f3694c2a219541dcbdb0269ce97e2d2e9d7
3627 Author: Jianpeng Ma <jianpeng.ma@intel.com>
3628 Date: Wed Aug 19 14:54:21 2015 +0800
3630 Objecter: Take RLocker when call is_active.
3632 Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
3633 (cherry picked from commit 89f0112e001a2561f9a5cd705898d43c8909501f)
3635 commit 684928883f5eaaec99a8cc8c43e81fae69f9fbfa
3636 Author: Kefu Chai <kchai@redhat.com>
3637 Date: Wed Sep 16 15:08:17 2015 +0800
3639 mon/PGMap: calc min_last_epoch_clean when decode
3642 Signed-off-by: Kefu Chai <kchai@redhat.com>
3643 (cherry picked from commit d0ac68bf1785b330f3202d924c2203ace9393fe6)
3645 commit 7d4b303b2d9649e2722a9b8f0be5c505776124df
3646 Author: Guang Yang <yguang@yahoo-inc.com>
3647 Date: Wed Sep 23 15:58:02 2015 +0000
3649 osd: print min_last_epoch_clean along with pg dump
3652 Signed-off-by: Guang Yang <yguang@yahoo-inc.com>
3653 (cherry picked from commit f5359f2314ce3e76ed446c01889dd23550c3ac36)
3655 commit d8ca88dd4b47b4956211d206b0d42112e5842d3f
3656 Author: Samuel Just <sjust@redhat.com>
3657 Date: Tue Sep 22 11:44:18 2015 -0700
3659 OSDService::agent_entry: don't use PG::operator<< without pg lock
3662 Signed-off-by: Samuel Just <sjust@redhat.com>
3663 (cherry picked from commit d422f2820d5beee15117e166546a6386120e0349)
3666 src/osd/OSD.cc: trivial resolution
3668 commit 397042a0cb2f4c628dd722192fee284359d2cbb3
3669 Author: Xinze Chi <xinze@xksy.com>
3670 Date: Wed Sep 23 19:26:00 2015 +0800
3672 filestore: fix peek_queue for OpSequencer
3675 Signed-off-by: Xinze Chi <xinze@xsky.com>
3676 (cherry picked from commit 5f7b3f5cde779f1300a35a04df2cdb1299f55472)
3678 commit d4e4d85c005ff68cb0461e9d7f0cbe5e6cf2f073
3679 Author: Sage Weil <sage@redhat.com>
3680 Date: Wed Sep 23 10:58:01 2015 -0400
3682 mon/Elector: do a trivial write on every election cycle
3684 Currently we already do a small write when the *first* election in
3685 a round happens (to update the election epoch). If the backend
3686 happens to fail while we are already in the midst of elections,
3687 however, we may continue to call elections without verifying we
3688 are still writeable.
3690 Signed-off-by: Sage Weil <sage@redhat.com>
3691 (cherry picked from commit ef909ccbdc303cce8a39edef255325127832ff16)
3693 commit 20a4c0c4831a9dcd49d8ad34b99369effbda9ab8
3694 Author: Sage Weil <sage@redhat.com>
3695 Date: Wed Sep 23 10:58:13 2015 -0400
3697 mon/MonitorDBStore: assert/crash if there is a write error
3699 Do this globally intead of relying on teh zillion mon callers to
3700 check the error code. There are no cases where we want to
3701 tolerate a commit failure.
3704 Signed-off-by: Sage Weil <sage@redhat.com>
3705 (cherry picked from commit 2fb7b1f0e33ada7c9a1be3de2f7708eb0760fcef)
3707 commit 3c1f7cbc45c4744eea3c821f0343596756e05ab1
3708 Author: Samuel Just <sjust@redhat.com>
3709 Date: Thu Sep 24 18:35:39 2015 -0700
3711 OSDMap: fill in known encode_features where possible
3713 Otherwise, if we get an incremental from hammer (struct_v = 6) we will
3714 encode the full map as if it were before CEPH_FEATURE_PGID64, which
3715 was actually pre-argonaut. Similarly, if struct_v >= 7, we know it
3716 was encoded with CEPH_FEATURE_OSDMAP_ENC.
3720 Signed-off-by: Samuel Just <sjust@redhat.com>
3721 (cherry picked from commit 04679c5451e353c966f6ed00b33fa97be8072a79)
3723 commit 93c523c65c1f0b6f9a76b1d71acf9e8df57c0fc9
3724 Author: Jason Dillaman <dillaman@redhat.com>
3725 Date: Tue Sep 8 15:47:37 2015 -0400
3727 librbd: diff_iterate needs to handle holes in parent images
3729 If a clone's parent image snapshot includes a discarded extent,
3730 this was previously causing an assert failure. Instead, ignore
3731 any discard holes in the parent image.
3735 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
3736 (cherry picked from commit 3ccc3bb4bd35e57209852d460633e371b4d004e2)
3739 src/librbd/DiffIterate.cc : file doesn't exist in hammer, logic is in internal.cc
3741 commit 359b68447393980bb7cdd4acc23582e8046b186c
3742 Author: Jason Dillaman <dillaman@redhat.com>
3743 Date: Mon Sep 28 10:25:07 2015 -0400
3745 tests: new test case for librbd diff_iterate over discard extents
3747 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
3748 (cherry picked from commit d5650c9cf85188efa73b279c8f4e4723fa475308)
3751 src/test/librbd/test_librbd.cc : diff_iterate2 doesn't exist in hammer
3753 commit 71a42f7c3aa55de688ef16ea7e392573fa5a7e62
3754 Author: Xinze Chi <xmdxcxz@gmail.com>
3755 Date: Mon Jun 8 16:54:08 2015 +0800
3757 bug fix: osd: avoid multi set osd_op.outdata in tier pool
3759 There are two read op on the same object for ec pool. First op read
3760 miss happend, calling do_proxy_read and promote_object, The second op only
3761 do_proxy_read. but before first op process_copy_chunk finish, the second op
3762 finish_proxy_read. first op receive reply from base pool
3763 first and then second received. so the second op set the field "outdata"
3764 in m->ops first. And then first op requeue_ops in process_copy_chunk,
3765 At last in do_osd_ops, it append outdata field.
3768 Signed-off-by: Xinze Chi <xmdxcxz@gmail.com>
3769 (cherry picked from commit 855ae1fd6e4557adba1cbd8ab532488b867cee2a)
3772 src/osd/ReplicatedPG.cc
3774 commit e8cce089102abd8ec1b8e916c6f2c4a5b339cbd3
3775 Author: Xinze Chi <xmdxcxz@gmail.com>
3776 Date: Wed Jul 29 16:28:33 2015 +0800
3778 bug fix: osd: requeue_scrub when kick_object_context_blocked
3780 when read miss in writeback cache pool, osd do_proxy_read first
3781 and maybe promote it. but in this case, the op is not added to
3782 waiting_for_blocked_object. pg scrub maybe block by this object
3783 (_range_available_for_scrub). so after promote it, we should
3787 Signed-off-by: Xinze Chi <xmdxcxz@gmail.com>
3788 (cherry picked from commit f77949fedce3449befd74efeb5270579f5085b16)
3791 src/osd/ReplicatedPG.cc
3792 in ReplicatedPG::kick_object_context_blocked, removed hobject_t::BitwiseComparator from map p
3794 commit 8ee93d5f661b303cafcb0134a0e9a1fce7b8e6b2
3795 Author: Samuel Just <sjust@redhat.com>
3796 Date: Thu Aug 27 11:08:33 2015 -0700
3798 PG::handle_advance_map: on_pool_change after handling the map change
3800 Otherwise, the is_active() checks in the hitset code can erroneously
3801 return true firing off repops stamped with the new epoch which then get
3802 cleared in the map change code. The filestore callbacks then pass the
3803 interval check and call into a destroyed repop structure.
3806 Backport: hammer,firefly
3807 Signed-off-by: Samuel Just <sjust@redhat.com>
3808 (cherry picked from commit 14e02bc90a463805f4c3e2de210892067a52514b)
3810 commit cecd68d06d4d114b0b735155da5c2bac736714c3
3811 Author: Kefu Chai <kchai@redhat.com>
3812 Date: Fri Aug 28 11:36:49 2015 +0800
3814 ceph_test_rados_api_aio: add a test for aio_sparse_read
3816 Signed-off-by: Kefu Chai <kchai@redhat.com>
3817 (cherry picked from commit 076bad955d374cbb37b77e2b0429f3c85f32abc0)
3819 commit 8849f9933fe0fc315da0981f9cd88ea108f686ed
3820 Author: Kefu Chai <kchai@redhat.com>
3821 Date: Thu Aug 27 22:57:16 2015 +0800
3823 ceph_test_rados_api_io: add tests for sparse_read
3825 Signed-off-by: Kefu Chai <kchai@redhat.com>
3826 (cherry picked from commit 4d4920610ebfcb516630ed15678979c9e9292f5a)
3829 src/test/librados/test.cc
3830 minor changes and remove C++11 stuff
3832 commit 64b22dcba316bf9ac5598c249e4fba47c157f588
3833 Author: Kefu Chai <kchai@redhat.com>
3834 Date: Wed Aug 26 15:41:13 2015 +0800
3836 ceph_test_rados: also send sparse_read in ReadOp
3838 Signed-off-by: Kefu Chai <kchai@redhat.com>
3839 (cherry picked from commit 5ae2e7a185b5f95753a09a89d7110fc38848a083)
3842 src/test/osd/RadosModel.h
3843 remove the piped-read test
3845 commit 6ca593026f81580ecbd02ad6140347f31086a87d
3846 Author: Kefu Chai <kchai@redhat.com>
3847 Date: Fri Aug 28 14:27:53 2015 +0800
3849 osd: should use ec_pool() when checking for an ecpool
3851 we were using pool.info.require_rollback() in do_osd_ops() when
3852 handling OP_SPARSE_READ to tell if a pool is an ecpool. should
3853 use pool.info.ec_pool() instead.
3855 Signed-off-by: Kefu Chai <kchai@redhat.com>
3856 (cherry picked from commit a5bfde69a9d14de67da1e3354173ec70ba089b37)
3858 commit 318049911445a3cc257db2b3beb3ab53e6f1d64e
3859 Author: Kefu Chai <kchai@redhat.com>
3860 Date: Thu Aug 27 22:57:49 2015 +0800
3862 osd: translate sparse_read to read for ecpool
3865 Signed-off-by: Kefu Chai <kchai@redhat.com>
3866 (cherry picked from commit 700d42ef1c82f5602249b96690ae881c1d259d54)
3868 commit d8ac5103e433f6649e28392b365d2e02b4ba1ffe
3869 Author: Jason Dillaman <dillaman@redhat.com>
3870 Date: Mon Aug 10 19:10:19 2015 -0400
3872 WorkQueue: add/remove_work_queue methods now thread safe
3874 These methods were not acquiring the ThreadPool lock when
3875 manipulating the work_queue collection. This was causing
3876 occasional crashes within librbd when opening and closing
3880 Backport: hammer, firefly
3881 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
3882 (cherry picked from commit 3e18449b01c1ab78d1bbfc1cf111aa9bdbef7b1f)
3884 commit e8d6d5a5f43b75f6aaec4c977221ad97d11e314c (refs/remotes/gh/wip-11786.hammer)
3885 Author: Joao Eduardo Luis <joao@suse.de>
3886 Date: Mon Sep 7 14:12:19 2015 +0100
3888 mon: MonitorDBStore: make get_next_key() work properly
3890 We introduced a significant bug with 2cc7aee, when we fixed issue #11786.
3891 Although that patch would fix the problem described in #11786, we
3892 managed to not increment the iterator upon returning the current key.
3893 This would have the iterator iterating over the same key, forever and
3896 Signed-off-by: Joao Eduardo Luis <joao@suse.de>
3897 (cherry picked from commit 70d31082fd3dc8c7857994104577f1a3631c678c)
3899 commit dba8b5b113f585586fcbbc0acf05543262486328
3900 Author: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
3901 Date: Tue May 19 13:44:21 2015 +0200
3903 tools: Don't delete, recreate and re-fill buffers in rados bench.
3905 Fixes the high CPU usage and corrects rados bench scores on fast SSDs
3906 and ramdisks/memstore.
3907 For bench run on SSD, on Intel(R) Xeon(R) CPU E5-2640 v2 @ 2.00GHz
3908 before this patch, times are:
3909 write: real 5m0.169s, user 2m33.565s, sys 4m39.791s
3910 seq: real 4m28.642s, user 1m35.250s, sys 6m42.948s
3911 rand: real 5m0.258s, user 1m19.656s, sys 6m47.145s
3914 write: real 5m1.162s, user 0m27.788s, sys 3m11.707s
3915 seq: real 5m1.149s, user 2m23.278s, sys 4m14.427s
3916 rand: real 5m1.021s, user 2m30.514s, sys 4m20.347s
3918 Bench run: rados -p ssd bench 300 write|seq|read --no-cleanup
3920 Note the increase in user time cpu on seq/read tests,
3921 along with decreased sys cpu time; this is because there's
3922 additional memcmp() that compares read objects with expected
3923 contents. With less time spent memory juggling, more time is
3924 spent performing more reads per second, increasing memcmp call
3925 count and increasing amount of user cpu time used.
3927 Signed-off-by: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
3928 (cherry picked from commit b894fc7acf7dee7f7ec8c5c280e7a6be41133328)
3931 src/common/obj_bencher.cc
3932 return -5 was replaced with return r
3934 commit fa72ecb509fa4ddd5f227b1e279eae390284a72c
3935 Author: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
3936 Date: Fri May 15 15:44:18 2015 +0200
3938 bufferlist: implement bufferlist::invalidate_crc()
3940 This function iterates over all bufferlist internal buffers and calls
3941 their invalidate_crc() method. Required for rados bench to work
3942 correctly, because it modifies buffers outside buffer api, invalidating
3943 internal CRC cache in the process - this method clears that cache, so
3944 another call for buffer::get_crc() to return correct checksum.
3946 Signed-off-by: Piotr Dałek <piotr.dalek@ts.fujitsu.com>
3947 (cherry picked from commit 55a6f9efbee041d041742810ca4fa8874b6191a7)
3949 commit 50c82f2095add5ad75200a62081fd2f5fea8d8d9
3950 Author: Owen Synge <osynge@suse.com>
3951 Date: Mon Jun 1 15:57:03 2015 +0200
3953 ceph.spec.in summary-ended-with-dot
3955 RPM spec files summary should not end with a dot.
3956 This was inconsistent across the ceph rpm packages
3957 and creates errors with rpm lint.
3959 Signed-off-by: Owen Synge <osynge@suse.com>
3960 (cherry picked from commit 042e8ff4dc71795d957bbfac12fade2bc0fc4674)
3962 commit 94699e724b45a9a6de84a943c5b834d2f20f85ef
3963 Author: Owen Synge <osynge@suse.com>
3964 Date: Mon Jun 1 14:58:31 2015 +0200
3966 ceph.spec.in libcephfs_jni1 has no %post and %postun
3968 /usr/lib64/libcephfs_jni.so.1.0.0 requires /sbin/ldconfig to be
3969 run after installing and after removing.
3971 Signed-off-by: Owen Synge <osynge@suse.com>
3972 (cherry picked from commit 0742d82b5561634a5f4ed73fc6cbe4d4f26b0d42)
3974 commit 7475a8fe3f31a46a949f051ded383cd6f396a0fa
3975 Author: Sage Weil <sage@redhat.com>
3976 Date: Tue Apr 21 16:08:09 2015 -0700
3978 logrotate.conf: prefer service over invoke-rc.d
3980 On trusty 14.04, service works but invoke-rc.d does not (but is present).
3983 Reported-by: Wim Heirman <wim@heirman.net>
3984 Signed-off-by: Sage Weil <sage@redhat.com>
3985 (cherry picked from commit 6063a21f9edbbf80e44d45ff52a9f4ba1a1cb2c9)
3987 commit 86f88dfce3c0e785d0ff48ec8892e53e6a53b080
3988 Author: John Spray <john.spray@redhat.com>
3989 Date: Thu Jul 23 10:15:56 2015 +0100
3991 mon: fix checks on mds add_data_pool
3993 Use same _check_pool helper as during fs creation.
3997 Signed-off-by: John Spray <john.spray@redhat.com>
3998 (cherry picked from commit 329741ec238d6af4b69bc838ba667be232741b3d)
4000 commit a65c39852775b17323fa4efee54bde12ce22a512
4001 Author: Kefu Chai <kchai@redhat.com>
4002 Date: Wed Jul 29 15:52:19 2015 +0800
4004 common/syncfs: fall back to sync(2) if syncfs(2) not available
4007 Signed-off-by: Kefu Chai <kchai@redhat.com>
4008 (cherry picked from commit 08210d6a85527b0759f6dc4411f742bc13a4939f)
4011 src/common/sync_filesystem.h
4012 - subtle difference in #if/#else/#endif structure
4014 commit e6eaa914d685f2a8f1f8cf3b01112a1b720e1358
4015 Author: Joao Eduardo Luis <joao@suse.de>
4016 Date: Wed May 27 22:28:49 2015 +0100
4018 mon: MonitorDBStore: get_next_key() only if prefix matches
4020 get_next_key() had a bug in which we would always return the first key
4021 from the iterator, regardless of whether its prefix had been specified
4026 Signed-off-by: Joao Eduardo Luis <joao@suse.de>
4027 (cherry picked from commit 2cc7aee1abe40453093306c8fef2312b650dff5d)