1 commit 87a7cec9ab11c677de2ab23a7668a77d2f5b955e (tag: refs/tags/v0.87.2, refs/remotes/gh/giant)
2 Author: Jenkins <jenkins@inktank.com>
3 Date: Fri Apr 24 12:31:27 2015 -0700
7 commit c1301e84aee0f399db85e2d37818a66147a0ce78
9 Author: Loic Dachary <loic@dachary.org>
10 Date: Tue Apr 7 21:08:24 2015 +0200
12 Merge pull request #4214 from dachary/wip-10430-giant
14 osd/osd_types.cc: 456: FAILED assert(m_seed < old_pg_num)
16 Reviewed-by: David Zafman <dzafman@redhat.com>
18 commit 1a13e1065829c59987c2f57a13eaa03de31df4ed
19 Merge: 1fb08d3 5f4e62f
20 Author: Loic Dachary <loic@dachary.org>
21 Date: Tue Apr 7 16:39:28 2015 +0200
23 Merge pull request #4258 from ceph/wip-10643-v2
25 mon: MDSMonitor: additional backports for #10643
27 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
28 Reviewed-by: Loic Dachary <ldachary@redhat.com>
30 commit 1fb08d3066b14b178a8912ffb3c9f50d2333738c
31 Merge: 90b37d9 7684ee2
32 Author: Josh Durgin <jdurgin@redhat.com>
33 Date: Thu Apr 2 08:55:48 2015 -0700
35 Merge pull request #4261 from ceph/wip-11303-giant
37 allow -L to disable lttng. Enable it by default
39 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
41 commit 7684ee25ac21810153a44bdc4fc00b36e39eb12f
42 Author: Alfredo Deza <adeza@redhat.com>
43 Date: Thu Apr 2 10:34:02 2015 -0400
45 allow -L to disable lttng. Enable it by default
47 Signed-off-by: Alfredo Deza <adeza@redhat.com>
49 commit 5f4e62f382767ee69e5b0c701b1a01d9e4132237
50 Author: Joao Eduardo Luis <joao@redhat.com>
51 Date: Fri Oct 17 19:08:20 2014 +0100
53 mon: MDSMonitor: wait for osdmon to be writable when requesting proposal
55 Otherwise we may end up requesting the osdmon to propose while it is
56 mid-proposal. We can't simply return EAGAIN to the user either because
57 then we would have to expect the user to be able to successfully race
58 with the whole cluster in finding a window in which 'mds fs new' command
59 would succeed -- which is not a realistic expectation. Having the
60 command to osdmon()->wait_for_writable() guarantees that the command
61 will be added to a queue and that we will, eventually, tend to it.
65 Signed-off-by: Joao Eduardo Luis <joao@redhat.com>
66 (cherry picked from commit 2ae1cba595d9c56a0a4c534b34fe25250e7eb2d5)
68 commit 257bd17db6470ca050403b1c8ff8daa94a4b80b5
69 Author: Joao Eduardo Luis <joao@redhat.com>
70 Date: Fri Oct 17 18:59:51 2014 +0100
72 mon: MDSMonitor: have management_command() returning int instead of bool
74 We can more easily differentiate between errors we get out of this
75 function, which makes the code a bit more versatile and readable.
77 Signed-off-by: Joao Eduardo Luis <joao@redhat.com>
78 (cherry picked from commit 0dd473cbad4f9ea403fe60badffdc6da4dd3aa3c)
80 commit 9e9c3c652339d85863af01cac621228f04eb4f18
81 Author: David Zafman <dzafman@redhat.com>
82 Date: Thu Oct 9 11:20:13 2014 -0700
84 osd: Get pgid ancestor from last_map when building past intervals
86 Fixed OSD::build_past_intervals_parallel() and PG::generate_past_intervals()
90 Signed-off-by: David Zafman <dzafman@redhat.com>
91 (cherry picked from commit 0c5b66da7a9ba516340d06d9e806beb9d1040d0e)
93 commit 90b37d9bdcc044e26f978632cd68f19ece82d19a
94 Merge: 2ccbc14 9f1f355
95 Author: Loic Dachary <loic-redhat@dachary.org>
96 Date: Thu Mar 26 07:58:14 2015 +0100
98 Merge pull request #4175 from wonzhq/objecter-timer-2
100 Objecter: failed assert(tick_event==NULL) at osdc/Objecter.cc
102 Reviewed-by: Loic Dachary <ldachary@redhat.com>
104 commit 9f1f35546e00e8f1ecbce0697d59b64f3537facf
105 Author: Zhiqiang Wang <zhiqiang.wang@intel.com>
106 Date: Wed Mar 25 16:32:44 2015 +0800
108 Objecter: failed assert(tick_event==NULL) at osdc/Objecter.cc
110 When the Objecter timer erases the tick_event from its events queue and
111 calls tick() to dispatch it, if the Objecter::rwlock is held by shutdown(),
112 it waits there to get the rwlock. However, inside the shutdown function,
113 it checks the tick_event and tries to cancel it. The cancel_event function
114 returns false since tick_event is already removed from the events queue. Thus
115 tick_event is not set to NULL in shutdown(). Later the tick function return
116 ealier and doesn't set tick_event to NULL as well. This leads to the assertion
119 This is a regression introduced by an incorrect conflict resolution when
120 d790833 was backported.
124 Signed-off-by: Zhiqiang Wang <zhiqiang.wang@intel.com>
126 commit 2ccbc14d17b54ea4fd4126cb04a7b83cd64c7f1e
127 Merge: 02f9cdb de4b087
128 Author: Loic Dachary <ldachary@redhat.com>
129 Date: Mon Mar 23 20:39:26 2015 +0100
131 Merge pull request #4127 from dzafman/wip-11176-giant
133 ceph-objectstore-tool: Output only unsupported features when incomatible
135 Reviewed-by: Loic Dachary <ldachary@redhat.com>
137 commit 02f9cdbf889071ca6fe3811d9b9a92a0b630fa55
138 Merge: 83bcc51 fc43d8c
139 Author: Loic Dachary <ldachary@redhat.com>
140 Date: Sun Mar 22 23:11:46 2015 +0100
142 Merge pull request #4097 from dachary/wip-10497-giant
144 librados: c api does not translate op flag
146 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
148 commit 83bcc516743e426c7a8c6f6401721bffbbec4fc0
149 Merge: ebab2bd d790833
150 Author: Loic Dachary <ldachary@redhat.com>
151 Date: Sun Mar 22 23:11:26 2015 +0100
153 Merge pull request #4096 from dachary/wip-9617-giant
155 objecter shutdown races with msg dispatch
157 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
159 commit ebab2bd5f36205d666673600624aaa3e5e06c405
160 Merge: e31c92d 970a797
161 Author: Loic Dachary <ldachary@redhat.com>
162 Date: Sun Mar 22 23:11:03 2015 +0100
164 Merge pull request #4095 from dachary/wip-9675-giant
166 splitting a pool doesn't start when rule_id != ruleset_id
168 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
170 commit e31c92d8dd814d276357e431ed87b93d75933f77
171 Merge: f0ec5e3 7653511
172 Author: Loic Dachary <ldachary@redhat.com>
173 Date: Sun Mar 22 23:10:42 2015 +0100
175 Merge pull request #4094 from dachary/wip-9891-giant
177 Assertion: os/DBObjectMap.cc: 1214: FAILED assert(0)
179 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
181 commit f0ec5e3b3ed58af65323bcc494e589935147aa45
182 Merge: dd7c15b 13b0147
183 Author: Loic Dachary <ldachary@redhat.com>
184 Date: Sun Mar 22 23:10:25 2015 +0100
186 Merge pull request #4093 from dachary/wip-9915-giant
188 osd: eviction logic reversed
190 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
192 commit dd7c15b2b24027a7cc5fa4bff21222c5a4606e60
193 Merge: 33b09e1 13b8364
194 Author: Loic Dachary <ldachary@redhat.com>
195 Date: Sun Mar 22 23:09:28 2015 +0100
197 Merge pull request #4092 from dachary/wip-9985-giant
199 osd: incorrect atime calculation
201 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
203 commit 33b09e18aa78252d94cbec8bf94ec97ed5bb1573
204 Merge: 950123e 5550cdd
205 Author: Loic Dachary <ldachary@redhat.com>
206 Date: Sun Mar 22 23:08:41 2015 +0100
208 Merge pull request #4091 from dachary/wip-9986-giant
210 objecter: map epoch skipping broken
212 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
214 commit 950123e4daa85562b2f52e0e12e0bae07f444095
215 Merge: abdc065 21f81b8
216 Author: Loic Dachary <ldachary@redhat.com>
217 Date: Sun Mar 22 23:08:16 2015 +0100
219 Merge pull request #4090 from dachary/wip-10059-giant
221 osd/ECBackend.cc: 876: FAILED assert(0)
223 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
225 commit abdc065acd6cafa6439e9c1724cc87de02352bd7
226 Merge: aee2825 1ccf583
227 Author: Loic Dachary <ldachary@redhat.com>
228 Date: Sun Mar 22 23:07:52 2015 +0100
230 Merge pull request #4089 from dachary/wip-10080-giant
232 Pipe::connect() cause osd crash when osd reconnect to its peer
234 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
236 commit aee28250fee66bca08d91e56ce47ec46c2e9fc24
237 Merge: 6582253 3e875ab
238 Author: Loic Dachary <ldachary@redhat.com>
239 Date: Sun Mar 22 23:07:20 2015 +0100
241 Merge pull request #4088 from dachary/wip-6003-giant
243 journal Unable to read past sequence 406 ...
245 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
247 commit 65822530ebfdea2feb9192c6eb6e3b8b9d60fe33
248 Merge: 4b20f2d 96a5c67
249 Author: Loic Dachary <ldachary@redhat.com>
250 Date: Sun Mar 22 23:06:51 2015 +0100
252 Merge pull request #4082 from dachary/wip-10106-giant
254 rgw acl response should start with <?xml version=1.0 ?>
256 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
258 commit 4b20f2d2d1ee52deed33617f000fa342ebce2e49
259 Merge: 7ff3a67 c7b02f5
260 Author: Loic Dachary <ldachary@redhat.com>
261 Date: Sun Mar 22 23:06:23 2015 +0100
263 Merge pull request #4078 from dachary/wip-11157-giant
265 doc,tests: force checkout of submodules
267 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
269 commit 7ff3a67c44ba3dc20a663a7dc6ba28c25714f063
270 Merge: 440e706 4d4eb9f
271 Author: Loic Dachary <ldachary@redhat.com>
272 Date: Sun Mar 22 23:05:20 2015 +0100
274 Merge pull request #4077 from dachary/wip-10150-giant
276 osd/ReplicatedPG.cc: 10853: FAILED assert(r >= 0) (in _scan_range)
278 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
280 commit 440e70607e7e3cd1d8ca33843c626109431caf8d
281 Merge: 66f639b 499d94f
282 Author: Loic Dachary <ldachary@redhat.com>
283 Date: Sun Mar 22 23:04:51 2015 +0100
285 Merge pull request #4076 from dachary/wip-10153-giant
287 Rados.shutdown() dies with Illegal instruction (core dumped)
289 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
291 commit 66f639b30ce6c74caae9397f20053761203f8e87
292 Merge: 43b45df b79852f
293 Author: Loic Dachary <ldachary@redhat.com>
294 Date: Sun Mar 22 23:04:25 2015 +0100
296 Merge pull request #4074 from dachary/wip-10220-giant
298 mon/Paxos.cc: 1033: FAILED assert(mon->is_leader())
300 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
302 commit 43b45dfacd598bab51fa06c5d0e2d0605d6e83d6
303 Merge: d282cfd b318e2f
304 Author: Loic Dachary <ldachary@redhat.com>
305 Date: Sun Mar 22 23:03:35 2015 +0100
307 Merge pull request #3548 from ceph/wip-10643
309 mon: MDSMonitor: missing backports for giant
311 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
312 Reviewed-by: Loic Dachary <ldachary@redhat.com>
314 commit d282cfd5ae3e65b74801cd27480ce8c0bd72d9a7
315 Merge: 3f3b981 681c99f
316 Author: Loic Dachary <ldachary@redhat.com>
317 Date: Sun Mar 22 22:12:58 2015 +0100
319 Merge pull request #4053 from dachary/wip-8011-giant
321 osd/ReplicatedPG.cc: 5244: FAILED assert(soid < scrubber.start || soid >= scrubber.end)
323 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
325 commit 3f3b98123cb46080068c1c73f4be41acbe18bd0d
326 Merge: 6919eb1 4427358
327 Author: Loic Dachary <ldachary@redhat.com>
328 Date: Sun Mar 22 22:12:40 2015 +0100
330 Merge pull request #4052 from dachary/wip-10844-giant
332 mon: caps validation should rely on EntityName instead of entity_name_t
334 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
336 commit 6919eb1684e34a6395963be6cc65215a51f5ba13
337 Merge: 8876585 1d4ffbe
338 Author: Loic Dachary <ldachary@redhat.com>
339 Date: Sun Mar 22 22:12:15 2015 +0100
341 Merge pull request #4050 from dachary/wip-10817-giant
343 WorkQueue: make timeout when calling WaitInterval configurable
345 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
347 commit 88765851bd56a6a526f6ab724920c8858ec4956c
348 Merge: 6da3171 92c352d
349 Author: Loic Dachary <ldachary@redhat.com>
350 Date: Sun Mar 22 22:11:43 2015 +0100
352 Merge pull request #4049 from dachary/wip-10787-giant
354 mon: OSDMonitor::map_cache is buggy, send_incremental is not conservative
356 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
357 Reviewed-by: Sage Weil <sage@redhat.com>
359 commit 6da3171f21b8e1a56fe941a5028f2ccfdccee18a
360 Merge: 5a6eefc 25fcaca
361 Author: Loic Dachary <ldachary@redhat.com>
362 Date: Sun Mar 22 22:09:28 2015 +0100
364 Merge pull request #4048 from dachary/wip-10770-giant
366 rgw: pending bucket index operations are not cancelled correctly
368 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
370 commit 5a6eefcb40c94da9a900d9893ecb6eaaf8fd0cea
371 Merge: c67a7a5 2858327
372 Author: Loic Dachary <ldachary@redhat.com>
373 Date: Sun Mar 22 22:09:06 2015 +0100
375 Merge pull request #4046 from dachary/wip-10723-giant
377 rados python binding leaks Ioctx objects
379 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
381 commit c67a7a52dc96f176431125921e36e4a2b8a30f1c
382 Merge: 41dcd2d d5b1b7e
383 Author: Loic Dachary <ldachary@redhat.com>
384 Date: Sun Mar 22 22:08:45 2015 +0100
386 Merge pull request #4044 from dachary/wip-10617-giant
388 osd: pgs for deleted pools don't finish getting removed if osd restarts
390 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
392 commit 41dcd2d9c307dd0b25d27bd6673943f3fdaaa28b
393 Merge: 42e7413 c3d998e
394 Author: Loic Dachary <ldachary@redhat.com>
395 Date: Sun Mar 22 22:08:20 2015 +0100
397 Merge pull request #4034 from dachary/wip-10475-giant
399 rgw: Swift API. Support for X-Remove-Container-Meta-{key} header.
401 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
403 commit 42e741339913990521c4509f4e266921f8dbe007
404 Merge: ffb76d1 fa8d454
405 Author: Loic Dachary <ldachary@redhat.com>
406 Date: Sun Mar 22 22:07:53 2015 +0100
408 Merge pull request #4033 from dachary/wip-10471-giant
410 rgw: index swift keys appropriately
412 Reviewed-by: Abhishek Lekshmanan <abhishek.lekshmanan@ril.com>
414 commit de4b08704172ac31b511dde50e5c11d58d811ca2
415 Author: David Zafman <dzafman@redhat.com>
416 Date: Fri Mar 20 16:57:40 2015 -0700
418 ceph-objectstore-tool: Output only unsupported features when incomatible
421 Backport: firefly, giant
423 Signed-off-by: David Zafman <dzafman@redhat.com>
424 (cherry picked from commit 5b23f5b5892b36fb7d06efc0d77e64a24ef6e8c9)
426 commit ffb76d16c5d9d65ac94d21b4e8fc1fdf86441977
427 Merge: fea29b1 61d6006
428 Author: John Spray <jcspray@gmail.com>
429 Date: Thu Mar 19 22:07:58 2015 +0000
431 Merge pull request #3971 from ceph/giant-11053
433 mds: fix assertion caused by system clock backwards
435 Reviewed-by: John Spray <john.spray@redhat.com>
437 commit c3d998e4c3c9d90b50c3f6c0af7d48785616e032
438 Author: Dmytro Iurchenko <diurchenko@mirantis.com>
439 Date: Tue Feb 3 17:54:38 2015 +0200
441 rgw: Swift API. Support for X-Remove-Container-Meta-{key} header.
444 Backport: hammer, firefly
445 Reported-by: Josh Durgin <jdurgin@redhat.com>
446 Signed-off-by: Dmytro Iurchenko <diurchenko@mirantis.com>
447 (cherry picked from commit f67bfa24fd6f69c2fcc0987eba8b6b426dd78320)
451 trivial merge: prototype of an unrelated function changed
453 s/is_object_op/!(s->object == NULL)/
455 commit fc43d8c2c54ec9e9cb6ef4d19cca695eb2fb3aab
456 Author: Matt Richards <mattjrichards@gmail.com>
457 Date: Thu Jan 8 13:16:17 2015 -0800
459 librados: Translate operation flags from C APIs
461 The operation flags in the public C API are a distinct enum
462 and need to be translated to Ceph OSD flags, like as happens in
463 the C++ API. It seems like the C enum and the C++ enum consciously
464 use the same values, so I reused the C++ translation function.
466 Signed-off-by: Matthew Richards <mattjrichards@gmail.com>
467 (cherry picked from commit 49d114f1fff90e5c0f206725a5eb82c0ba329376)
469 commit d790833cb84d6f6349146e4f9abdcdffb4db2ee0
470 Author: Josh Durgin <josh.durgin@inktank.com>
471 Date: Mon Sep 29 18:17:29 2014 -0700
473 Objecter: check the 'initialized' atomic_t safely
475 shutdown() resets initialized to 0, but we can still receive messages
476 after this point, so fix message handlers to skip messages in this
477 case instead of asserting.
479 Also read initialized while holding Objecter::rwlock to avoid races
480 where e.g. handle_osd_map() checks initialized -> 1, continues,
481 shutdown() is called, sets initialized to 0, then handle_osd_map()
482 goes about its business and calls op_submit(), which would fail the
483 assert(initialized.read()) check. Similar races existed in other
484 message handlers which change Objecter state.
486 The Objecter is not destroyed until after its Messenger in
487 the MDS, OSD, and librados, so this should be safe.
491 Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
492 (cherry picked from commit e506f896a9217324ab7a7865989f4454562aed5f)
496 context changed: Objecter::tick() did not have
497 assert(initialized.read())
499 commit ce436a33e0f720ea4b8cf0363bcac1126be3c28b
500 Author: Josh Durgin <josh.durgin@inktank.com>
501 Date: Mon Sep 29 18:12:50 2014 -0700
503 Objecter: init with a constant of the correct type
507 Signed-off-by: Josh Durgin <josh.durgin@inktank.com>
508 (cherry picked from commit 1feba200aae7d9a042cda705c3de8fba2fc82331)
510 commit 970a79753dc52d82d2abf29ffe6f88adac678eb0
511 Author: Xiaoxi Chen <xiaoxi.chen@intel.com>
512 Date: Wed Aug 20 15:35:44 2014 +0800
514 CrushWrapper: pick a ruleset same as rule_id
516 Originally in the add_simple_ruleset funtion, the ruleset_id
517 is not reused but rule_id is reused. So after some add/remove
518 against rules, the newly created rule likely to have
521 We dont want this happen because we are trying to hold the constraint
522 that ruleset == rule_id.
524 Signed-off-by: Xiaoxi Chen <xiaoxi.chen@intel.com>
525 (cherry picked from commit 78e84f34da83abf5a62ae97bb84ab70774b164a6)
527 commit 76535116823f02f0392226e5725fbfef14c277ba
528 Author: Samuel Just <sjust@redhat.com>
529 Date: Fri Feb 20 13:43:46 2015 -0800
531 DBObjectMap: lock header_lock on sync()
533 Otherwise, we can race with another thread updating state.seq
534 resulting in the old, smaller value getting persisted. If there
535 is a crash at that time, we will reuse a sequence number, resulting
536 in an inconsistent node tree and bug #9891.
539 Backport: giant, firefly, dumpling
540 Signed-off-by: Samuel Just <sjust@redhat.com>
541 (cherry picked from commit 2b63dd25fc1c73fa42e52e9ea4ab5a45dd9422a0)
544 src/os/DBObjectMap.cc
545 because we have state.v = 1; instead of state.v = 2;
547 commit 13b01473c5c1a116d6cd6acbbb6cbe08ee6e9433
548 Author: Zhiqiang Wang <zhiqiang.wang@intel.com>
549 Date: Tue Oct 28 09:37:11 2014 +0800
551 osd: cache tiering: fix the atime logic of the eviction
553 Reported-by: Xinze Chi <xmdxcxz@gmail.com>
554 Signed-off-by: Zhiqiang Wang <zhiqiang.wang@intel.com>
555 (cherry picked from commit 622c5ac41707069ef8db92cb67c9185acf125d40)
557 commit 13b8364dc869b2eefcb6646cff1e18c31126ce7d
558 Author: Sage Weil <sage@redhat.com>
559 Date: Fri Oct 31 19:33:59 2014 -0700
561 osd/ReplicatedPG: fix compile error
563 From 1fef4c3d541cba360738437420ebfa2447d5802e.
565 Signed-off-by: Sage Weil <sage@redhat.com>
566 (cherry picked from commit 4a9ad7dc2da6f4fa6a64235776a3f1d2799aef60)
568 commit 55541b90db0a6d6c26c1fae6a4958fa5a320f82c
569 Author: Xinze Chi <xmdxcxz@gmail.com>
570 Date: Wed Oct 29 07:11:11 2014 +0000
572 Get the currently atime of the object in cache pool for eviction
574 Because if there are mutiple atime in agent_state for the same object, we should use the recently one.
576 Signed-off-by: Xinze Chi <xmdxcxz@gmail.com>
577 (cherry picked from commit 1fef4c3d541cba360738437420ebfa2447d5802e)
579 commit 5550cdd876db913b152edad17a4de16bded31627
580 Author: Ding Dinghua <dingdinghua85@gmail.com>
581 Date: Thu Oct 30 14:58:42 2014 +0800
583 osdc/Objecter: Fix a bug of dead looping in Objecter::handle_osd_map
585 If current map epoch is less than oldest epoch, current map epoch
586 should step up to oldest epoch.
589 Signed-off-by: Ding Dinghua <dingdinghua85@gmail.com>
590 (cherry picked from commit e0166a23c2cf655bfb4cf873be021a14d9b9be27)
592 commit 7c5056f5cf77065e0e5a920f525f0a3be6b4b299
593 Author: Ding Dinghua <dingdinghua85@gmail.com>
594 Date: Thu Oct 30 14:58:05 2014 +0800
596 osdc/Objecter: e shouldn't be zero in Objecter::handle_osd_map
598 Signed-off-by: Ding Dinghua <dingdinghua85@gmail.com>
599 (cherry picked from commit 31c584c8ba022cd44fe2872d221f3026618cefab)
601 commit 21f81b85de652aede51e88b87bdff71f2e411da3
602 Author: Samuel Just <sjust@redhat.com>
603 Date: Wed Nov 19 08:20:16 2014 -0800
605 PG: always clear_primary_state on new interval, but only clear pg temp if not primary
607 Signed-off-by: Samuel Just <sjust@redhat.com>
608 (cherry picked from commit f692bfe076b8ddb679c6d1a6ea78cc47f0876326)
610 commit 0c3f7637d8cf2146a4268330d8c0506ad38c354d
611 Author: Samuel Just <sjust@redhat.com>
612 Date: Fri Nov 14 15:44:20 2014 -0800
614 PG: always clear_primary_state when leaving Primary
616 Otherwise, entries from the log collection process might leak into the next
617 epoch, where we might end up choosing a different authoritative log. In this
618 case, it resulted in us not rolling back to log entries on one of the replicas
619 prior to trying to recover from an affected object due to the peer_missing not
623 Backport: giant, firefly, dumpling
624 Signed-off-by: Samuel Just <sjust@redhat.com>
625 (cherry picked from commit c87bde64dfccb5d6ee2877cc74c66fc064b1bcd7)
627 commit 1ccf58355d0796172935938572cf68ceb31a6801
628 Author: Greg Farnum <gfarnum@redhat.com>
629 Date: Tue Dec 2 15:17:57 2014 -0800
631 SimpleMessenger: allow RESETSESSION whenever we forget an endpoint
633 In the past (e229f8451d37913225c49481b2ce2896ca6788a2) we decided to disable
634 reset of lossless Pipes, because lossless peers resetting caused trouble and
635 they can't forget about each other. But they actually can: if mark_down()
638 I can't figure out how else we could forget about a remote endpoint, so I think
639 it's okay if we tell them we reset in order to clean up state. That's desirable
640 so that we don't get into strange situations with out-of-whack counters.
643 Backport: giant, firefly, dumpling
645 Signed-off-by: Greg Farnum <gfarnum@redhat.com>
646 (cherry picked from commit 8cd1fdd7a778eb84cb4d7161f73bc621cc394261)
648 commit 3e875ab108de8d2aa3717f76a3fe48ede286abb7
649 Author: Samuel Just <sjust@redhat.com>
650 Date: Fri Feb 6 09:52:29 2015 -0800
652 FileJournal: fix journalq population in do_read_entry()
655 Backport: dumpling, firefly, giant
656 Signed-off-by: Samuel Just <sjust@redhat.com>
657 (cherry picked from commit bae1f3eaa09c4747b8bfc6fb5dc673aa6989b695)
660 src/os/FileJournal.cc
661 because reinterpret_cast was added near two hunks after firefly
663 commit 96a5c67121dde0d4d4cd13793bb131414b64cc28
664 Author: Yehuda Sadeh <yehuda@redhat.com>
665 Date: Fri Jan 30 18:42:40 2015 -0800
667 rgw: flush xml header on get acl request
670 Backport: firefly, giant
672 dump_start() updates the formatter with the appropriate prefix, however,
673 we never flushed the formatter.
675 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
676 (cherry picked from commit eb45f861343162e018968b8c56693a8c6f5b2cab)
678 commit c7b02f5a7347ea8688c5214f85bb9f612925a586
679 Author: Loic Dachary <ldachary@redhat.com>
680 Date: Thu Mar 19 00:32:39 2015 +0100
682 doc,tests: force checkout of submodules
684 When updating submodules, always checkout even if the HEAD is the
685 desired commit hash (update --force) to avoid the following:
687 * a directory gmock exists in hammer
688 * a submodule gmock replaces the directory gmock in master
689 * checkout master + submodule update : gmock/.git is created
690 * checkout hammer : the gmock directory still contains the .git from
691 master because it did not exist at the time and checkout won't
692 remove untracked directories
693 * checkout master + submodule update : git rev-parse HEAD is
694 at the desired commit although the content of the gmock directory
697 http://tracker.ceph.com/issues/11157 Fixes: #11157
699 Signed-off-by: Loic Dachary <ldachary@redhat.com>
701 commit 4d4eb9faf1871c4469b78a7ee75d527ce5cc67ad
702 Author: Samuel Just <sjust@redhat.com>
703 Date: Thu Dec 11 13:05:54 2014 -0800
705 ReplicatedPG::scan_range: an object can disappear between the list and the attr get
707 The first item in the range is often last_backfill, upon which writes
708 can be occuring. It's trimmed off on the primary side anyway.
711 Backport: dumpling, firefly, giant
712 Signed-off-by: Samuel Just <sjust@redhat.com>
713 (cherry picked from commit dce6f288ad541fe7f0ef8374301cd712dd3bfa39)
715 commit 499d94f573e78f6545bd5a12ade6b5ba5a75bbca
716 Author: Federico Simoncelli <fsimonce@redhat.com>
717 Date: Sat Nov 15 14:14:04 2014 +0000
719 common: do not unlock rwlock on destruction
721 According to pthread_rwlock_unlock(3p):
723 Results are undefined if the read-write lock rwlock is not held
724 by the calling thread.
728 https://sourceware.org/bugzilla/show_bug.cgi?id=17561
730 Calling pthread_rwlock_unlock on an rwlock which is not locked
733 calling pthread_rwlock_unlock on RWLock destruction could cause
734 an unknown behavior for two reasons:
736 - the lock is acquired by another thread (undefined)
737 - the lock is not acquired (undefined)
739 Moreover since glibc-2.20 calling pthread_rwlock_unlock on a
740 rwlock that is not locked results in a SIGILL that kills the
743 This patch removes the pthread_rwlock_unlock call on destruction
744 and replaces it with an assertion to check that the RWLock is
747 Any code that relied on the implicit release is now going to
748 break the assertion, e.g.:
753 } // implicit release, wrong.
755 Signed-off-by: Federico Simoncelli <fsimonce@redhat.com>
756 (cherry picked from commit cf2104d4d991361c53f6e2fea93b69de10cd654b)
758 commit b79852f3ab0643bbb3f6b71a56b15e5a4b5fc1f5
759 Author: Joao Eduardo Luis <joao@redhat.com>
760 Date: Wed Dec 10 17:46:35 2014 +0000
762 mon: Paxos: reset accept timeout before submiting work to the store
764 Otherwise we may trigger the timeout while waiting for the work to be
765 committed to the store -- and it would only take the write to take a bit
766 longer than 10 seconds (default accept timeout).
768 We do wait for the work to be properly committed to the store before
769 extending the lease though.
773 Signed-off-by: Joao Eduardo Luis <joao@redhat.com>
774 (cherry picked from commit 18534615f184ba56b441fd1d4242eb06debdfe13)
776 commit e997c9fed5feb3e877dfe07ffac1327b85d09ea2
777 Author: Joao Eduardo Luis <joao@redhat.com>
778 Date: Tue Dec 9 17:35:47 2014 +0000
780 mon: MonitorDBStore: allow randomly injecting random delays on writes
782 Adds two new config options:
784 mon_inject_transaction_delay_probability : DOUBLE (0.0-1.0, default: 0.0)
785 mon_inject_transaction_delay_max : DOUBLE (seconds, default: 10.0)
787 If probability is set to a value greater than 0, just before applying
788 the transaction, the store will decide whether to inject a delay,
789 randomly choosing a value between 0 and the max.
791 Signed-off-by: Joao Eduardo Luis <joao@redhat.com>
792 (cherry picked from commit beaa04e4119765d5775a6c48fd072dd95c984e3b)
794 commit 1d4ffbe1d233de555c1ecb0a33eebe2391b29f33
795 Author: Samuel Just <sjust@redhat.com>
796 Date: Mon Feb 9 17:41:19 2015 -0800
798 ShardedThreadPool: make wait timeout on empty queue configurable
802 Signed-off-by: Samuel Just <sjust@redhat.com>
803 (cherry picked from commit 7002f934e6664daa995ca0629c0ea3bae1c6bddf)
805 commit 292c4339c064968c2aa05eec701fbc2a8d82dab0
806 Author: Samuel Just <sjust@redhat.com>
807 Date: Mon Feb 9 17:11:38 2015 -0800
809 WorkQueue: make wait timeout on empty queue configurable
812 Backport: giant, firefly, dumpling
813 Signed-off-by: Samuel Just <sjust@redhat.com>
814 (cherry picked from commit 5aa6f910843e98a05bfcabe6f29d612cf335edbf)
816 commit fea29b1bcbd17b3d1f642398ec70dbe258bbc98f
817 Author: Samuel Just <sjust@redhat.com>
818 Date: Thu Nov 20 15:15:08 2014 -0800
820 PGLog: include rollback_info_trimmed_to in (read|write)_log
823 Backport: firefly, giant
824 Signed-off-by: Samuel Just <sjust@redhat.com>
825 (cherry picked from commit 1fe8b846641486cc294fe7e1d2450132c38d2dba)
827 commit 4427358bb556d902b01df27fd097cc3eefa561da
828 Author: Joao Eduardo Luis <joao@redhat.com>
829 Date: Wed Feb 11 23:36:01 2015 +0000
831 mon: MonCap: take EntityName instead when expanding profiles
833 entity_name_t is tightly coupled to the messenger, while EntityName is
834 tied to auth. When expanding profiles we want to tie the profile
835 expansion to the entity that was authenticated. Otherwise we may incur
836 in weird behavior such as having caps validation failing because a given
837 client messenger inst does not match the auth entity it used.
841 ceph --name osd.0 config-key exists foo daemon-private/osd.X/foo
843 has entity_name_t 'client.12345' and EntityName 'osd.0'. Using
844 entity_name_t during profile expansion would not allow the client access
845 to daemon-private/osd.X/foo (client.12345 != osd.X).
848 Backport: firefly,giant
850 Signed-off-by: Joao Eduardo Luis <joao@redhat.com>
851 (cherry picked from commit 87544f68b88fb3dd17c519de3119a9ad9ab21dfb)
853 commit 8902279dd4b52516d59db712bd59e8d9372611a6
854 Author: Joao Eduardo Luis <joao@redhat.com>
855 Date: Fri Nov 14 21:03:54 2014 +0000
857 mon: Monitor: stash auth entity name in session
861 Signed-off-by: Joao Eduardo Luis <joao@redhat.com>
862 (cherry picked from commit ca8e1efc0be9bffcfbdce5593526d257aa498062)
864 commit 681c99fe12cfa6318f8cf06f4e825805635bcc07
865 Author: Samuel Just <sjust@redhat.com>
866 Date: Thu Nov 20 14:27:39 2014 -0800
868 ReplicatedPG: fail a non-blocking flush if the object is being scrubbed
871 Backport: firefly, giant
872 Signed-off-by: Samuel Just <sjust@redhat.com>
873 (cherry picked from commit 9b26de3f3653d38dcdfc5b97874089f19d2a59d7)
875 commit 1f58a0adc3b23785fac00083d721b62f6a4c44a1
876 Merge: 9fee7ba fe7d4ca
877 Author: Sage Weil <sage@redhat.com>
878 Date: Tue Mar 17 10:52:01 2015 -0700
880 Merge pull request #4042 from dachary/wip-10546-giant
882 ceph time check start round bug in monitor.cc
884 commit 9fee7ba9d7343bda29bc5113f8db5cdd9c09d71a
885 Merge: ca635ce e7af52a
886 Author: Sage Weil <sage@redhat.com>
887 Date: Tue Mar 17 10:50:26 2015 -0700
889 Merge pull request #4047 from dachary/wip-10762-giant
891 mon: osd gets marked down twice
893 commit ca635ce3eb21135b08613b97ce0c56977de44351
894 Merge: 9094185 5771f57
895 Author: Sage Weil <sage@redhat.com>
896 Date: Tue Mar 17 10:49:53 2015 -0700
898 Merge pull request #4041 from dachary/wip-10512-giant
900 osd: cancel_flush requeues blocked events after blocking event
902 commit 90941850ceb2c080ac72c67fed991c951b13f449
903 Merge: bc1b9d2 5fca232
904 Author: Sage Weil <sage@redhat.com>
905 Date: Tue Mar 17 10:47:26 2015 -0700
907 Merge pull request #4031 from dachary/wip-10353-giant
909 crush: set_choose_tries = 100 for erasure code rulesets
911 commit bc1b9d22d29a74e6853834fb0430e8803710bbbf
912 Merge: bd69cfc 30a1744
913 Author: Sage Weil <sage@redhat.com>
914 Date: Tue Mar 17 10:47:08 2015 -0700
916 Merge pull request #4029 from dachary/wip-9910-giant
918 msg/Pipe: discard delay queue before incoming queue
920 Reviewed-by: Greg Farnum <gfarnum@redhat.com>
922 commit bd69cfcbd58103843fbc42c4e55d43c2bafe54ff
923 Merge: 30f7df2 4bd5237
924 Author: Sage Weil <sage@redhat.com>
925 Date: Tue Mar 17 10:44:53 2015 -0700
927 Merge pull request #4030 from dachary/wip-10351-giant
929 mount.ceph: avoid spurious error message
931 Reviewed-by: John Spray <john.spray@redhat.com>
933 commit 30f7df2b110f73477354fbfa607b4777317dddfa
934 Merge: e5a50c3 8307318
935 Author: Sage Weil <sage@redhat.com>
936 Date: Tue Mar 17 10:44:00 2015 -0700
938 Merge pull request #4028 from dachary/wip-10259-giant
940 osd_types: op_queue_age_hist and fs_perf_stat should be in osd_stat_t::o...
942 commit e5a50c3ae42c8b9d16a19a13e10fc35b91149e73
943 Merge: ba1d55d a5a76f8
944 Author: Sage Weil <sage@redhat.com>
945 Date: Tue Mar 17 10:42:10 2015 -0700
947 Merge pull request #4027 from dachary/wip-10257-giant
949 mon: PGMonitor: several stats output error fixes
951 commit ba1d55de1e92a0143702e9e3fbbb1820db17d6c3
952 Merge: 530fce4 640986f
953 Author: Sage Weil <sage@redhat.com>
954 Date: Tue Mar 17 10:41:56 2015 -0700
956 Merge pull request #3998 from dzafman/wip-10677
958 Fix ceph command manpage to match ceph -h (giant)
960 Reviewed-by: Xinxin Shu <xinxin.shu@intel.com>
962 commit 530fce4d7cf51f4a4ac1e70ebe20b2795a30010e
963 Merge: a2a3dad 8a5a3c7
964 Author: Sage Weil <sage@redhat.com>
965 Date: Tue Mar 17 10:40:02 2015 -0700
967 Merge pull request #3921 from sponce/wip-11078-giant
969 Fix libstriprados::stat, use strtoll insdead of strtol
971 commit a2a3dad8ec00434b99d6d538615bf8ff9eada217
972 Merge: 0f57188 984df3b
973 Author: Sage Weil <sage@redhat.com>
974 Date: Tue Mar 17 10:36:28 2015 -0700
976 Merge pull request #3819 from tchaikov/giant-pg-leak-10421
978 osd: fix PG leak in SnapTrimWQ._clear()
980 commit 0f5718822793074ffab45124afb3bf3160e92004
981 Merge: 776c82c c5c6fcd
982 Author: Sage Weil <sage@redhat.com>
983 Date: Tue Mar 17 10:35:37 2015 -0700
985 Merge pull request #3771 from ceph/wip-10883-giant
987 osd: Fix FileJournal wrap to get header out first
989 commit 776c82caac36de14a5f895688f81c88a883fca3e
990 Merge: 938e036 37f196e
991 Author: Sage Weil <sage@redhat.com>
992 Date: Tue Mar 17 10:35:29 2015 -0700
994 Merge pull request #3637 from sponce/wip-10758-giant
996 Backport of pull request 3633 to giant : Fixed write_full behavior in libradosstriper
998 commit 92c352d047c84035478e71828e32f554d9f507fc
999 Author: Sage Weil <sage@redhat.com>
1000 Date: Thu Feb 12 13:49:50 2015 -0800
1002 mon/OSDMonitor: do not trust small values in osd epoch cache
1004 If the epoch cache says the osd has epoch 100 and the osd is asking for
1005 epoch 200+, do not send it 100+.
1008 Backport: giant, firefly
1009 Signed-off-by: Sage Weil <sage@redhat.com>
1010 (cherry picked from commit a5759e9b97107488a8508f36adf9ca1aba3fae07)
1012 commit 25fcaca19c76765284787c61584d25efd105a0ca
1013 Author: Yehuda Sadeh <yehuda@redhat.com>
1014 Date: Thu Feb 5 09:33:26 2015 -0800
1016 rgw: send appropriate op to cancel bucket index pending operation
1019 Backport: firefly, giant
1021 Reported-by: baijiaruo <baijiaruo@126.com>
1022 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
1023 (cherry picked from commit dfee96e3aebcaeef18c721ab73f0460eba69f1c7)
1026 src/rgw/rgw_rados.cc
1027 resolved by manual s/ADD/CANCEL/
1029 commit e7af52a147cc6e1a578d51193a2d5a425cee8a20
1030 Author: Sage Weil <sage@redhat.com>
1031 Date: Thu Feb 5 03:07:50 2015 -0800
1033 mon: ignore osd failures from before up_from
1035 If the failure was generated for an instance of the OSD prior to when
1036 it came up, ignore it.
1038 This probably causes a fair bit of unnecessary flapping in the wild...
1040 Backport: giant, firefly
1042 Reported-by: Dan van der Ster <dan@vanderster.com>
1043 Signed-off-by: Sage Weil <sage@redhat.com>
1044 (cherry picked from commit 400ac237d35d0d1d53f240fea87e8483c0e2a7f5)
1046 commit 28583276254ba6121b93688a6c6d8fab6588c864
1047 Author: Josh Durgin <jdurgin@redhat.com>
1048 Date: Mon Feb 9 20:50:23 2015 -0800
1050 rados.py: keep reference to python callbacks
1052 If we don't keep a reference to these, the librados aio calls will
1053 segfault since the python-level callbacks will have been garbage
1054 collected. Passing them to aio_create_completion() does not take a
1055 reference to them. Keep a reference in the python Completion object
1056 associated with the request, since they need the same lifetime.
1058 This fixes a regression from 60b019f69aa0e39d276c669698c92fc890599f50.
1061 Backport: dumpling, firefly, giant
1062 Signed-off-by: Josh Durgin <jdurgin@redhat.com>
1063 (cherry picked from commit 36d37aadbbbece28d70e827511f1a473d851463d)
1064 (cherry picked from commit 5f1245e131e33a98572408c8223deed2c7cf7b75)
1066 commit 1b5fb51a995d5d672a46176240d5887354c12967
1067 Author: Billy Olsen <billy.olsen@gmail.com>
1068 Date: Mon Feb 2 16:24:59 2015 -0700
1070 Fix memory leak in python rados bindings
1072 A circular reference was inadvertently created when using the
1073 CFUNCTYPE binding for callbacks for the asynchronous i/o callbacks.
1074 This commit refactors the usage of the callbacks such that the
1075 Ioctx object does not have a class reference to the callbacks.
1078 Backport: giant, firefly, dumpling
1079 Signed-off-by: Billy Olsen <billy.olsen@gmail.com>
1080 Reviewed-by: Dan Mick <dmick@redhat.com>
1081 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
1082 (cherry picked from commit 60b019f69aa0e39d276c669698c92fc890599f50)
1084 commit d5b1b7e52242e082820e0cdd76dc59e9aea7f284
1085 Author: Sage Weil <sage@redhat.com>
1086 Date: Fri Jan 23 10:47:44 2015 -0800
1088 osd: do not ignore deleted pgs on startup
1090 These need to get instantiated so that we can complete the removal process.
1093 Signed-off-by: Sage Weil <sage@redhat.com>
1094 (cherry picked from commit 879fd0c192f5d3c6afd36c2df359806ea95827b8)
1096 commit fe7d4ca17dacf4419be7bdc50010115cee36517b
1097 Author: Joao Eduardo Luis <joao@redhat.com>
1098 Date: Fri Jan 30 11:37:28 2015 +0000
1100 mon: Monitor: fix timecheck rounds period
1103 Backports: dumpling?,firefly,giant
1105 Signed-off-by: Joao Eduardo Luis <joao@redhat.com>
1106 (cherry picked from commit 2e749599ac6e1060cf553b521761a93fafbf65bb)
1108 commit 5771f57af731e61bf70f630cf85f5b94dcfd1edb
1109 Author: Sage Weil <sage@redhat.com>
1110 Date: Sun Jan 11 17:28:04 2015 -0800
1112 osd: requeue blocked op before flush it was blocked on
1114 If we have request A (say, cache-flush) that blocks things, and then
1115 request B that gets blocked on it, and we have an interval change, then we
1116 need to requeue B first, then A, so that the resulting queue will keep
1117 A before B and preserve the order.
1119 This was observed on this firefly run:
1121 ubuntu@teuthology:/a/sage-2015-01-09_21:43:43-rados-firefly-distro-basic-multi/694675
1123 Backport: giant, firefly
1125 Signed-off-by: Sage Weil <sage@redhat.com>
1126 (cherry picked from commit 11bdfb4131ecac16d4a364d651c6cf5d1d28c702)
1128 commit fa8d45423e22f7b916a345024a9f70283fc3097d
1129 Author: Yehuda Sadeh <yehuda@redhat.com>
1130 Date: Wed Jan 7 13:56:14 2015 -0800
1132 rgw: index swift keys appropriately
1135 Backport: firefly, giant
1137 We need to index the swift keys by the full uid:subuser when decoding
1138 the json representation, to keep it in line with how we store it when
1139 creating it through other mechanism.
1141 Reported-by: hemant burman <hemant.burman@gmail.com>
1142 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
1143 (cherry picked from commit 478629bd2f3f32afbe6e93eaebb8a8fa01af356f)
1145 commit 5fca232080c0bde50b5c65c60c133bf81311e8d3
1146 Author: Loic Dachary <ldachary@redhat.com>
1147 Date: Wed Dec 17 16:06:55 2014 +0100
1149 crush: set_choose_tries = 100 for erasure code rulesets
1151 It is common for people to try to map 9 OSDs out of a 9 OSDs total ceph
1152 cluster. The default tries (50) will frequently lead to bad mappings for
1153 this use case. Changing it to 100 makes no significant CPU performance
1154 difference, as tested manually by running crushtool on one million
1157 http://tracker.ceph.com/issues/10353 Fixes: #10353
1159 Signed-off-by: Loic Dachary <ldachary@redhat.com>
1160 (cherry picked from commit 2f87ac807f3cc7ac55d9677d2051645bf5396a62)
1162 commit 4bd523744cefb4a5e6b0d4440b9829b0e19dc012
1163 Author: Yan, Zheng <zyan@redhat.com>
1164 Date: Sat Jan 3 15:29:29 2015 +0800
1166 mount.ceph: avoid spurious error message
1168 /etc/mtab in most modern distributions is a symbol link to
1172 Signed-off-by: Yan, Zheng <zyan@redhat.com>
1173 (cherry picked from commit bdd0e3c4bda97fe18487a58dd173a7dff752e1a2)
1175 commit 30a17441f5030ac5c1227a9672811f88e580b4b2
1176 Author: Sage Weil <sage@redhat.com>
1177 Date: Wed Oct 29 14:45:11 2014 -0700
1179 msg/Pipe: discard delay queue before incoming queue
1181 Shutdown the delayed delivery before the incoming queue in case the
1182 DelayedDelivery thread is busy queuing messages.
1185 Signed-off-by: Sage Weil <sage@redhat.com>
1186 Reviewed-by: Greg Farnum <greg@inktank.com>
1187 (cherry picked from commit f7431cc3c25878057482007beb874c9d4473883e)
1189 commit 8307318cdf9487dde3231aa4a00fca6079379738
1190 Author: Samuel Just <sjust@redhat.com>
1191 Date: Fri Dec 5 15:29:52 2014 -0800
1193 osd_types: op_queue_age_hist and fs_perf_stat should be in osd_stat_t::operator==
1196 Backport: giant, firefly, dumpling
1197 Signed-off-by: Samuel Just <sjust@redhat.com>
1198 (cherry picked from commit 1ac17c0a662e6079c2c57edde2b4dc947f547f57)
1200 commit a5a76f81ae9242a54d6ef150fa0ac32b31e90c25
1201 Author: Joao Eduardo Luis <joao@redhat.com>
1202 Date: Mon Jan 19 18:49:15 2015 +0000
1204 mon: PGMonitor: skip zeroed osd stats on get_rule_avail()
1208 Signed-off-by: Joao Eduardo Luis <joao@redhat.com>
1209 (cherry picked from commit b311e7c36273efae39aa2602c1f8bd90d39e5975)
1211 commit 95ec68505ce4a26ea707716791cfb1e46a75ed25
1212 Author: Joao Eduardo Luis <joao@redhat.com>
1213 Date: Fri Jan 16 18:13:05 2015 +0000
1215 mon: PGMonitor: available size 0 if no osds on pool's ruleset
1217 get_rule_avail() may return < 0, which we were using blindly assuming it
1218 would always return an unsigned value. We would end up with weird
1219 values if the ruleset had no osds.
1221 Signed-off-by: Joao Eduardo Luis <joao@redhat.com>
1222 (cherry picked from commit 8be6a6ab2aa5a000a39c73a98b11a0ab32fffa1c)
1224 commit 1c480b39acc184018d976beddc55187f86dfff1a
1225 Author: Joao Eduardo Luis <joao@redhat.com>
1226 Date: Fri Jan 16 18:12:42 2015 +0000
1228 mon: PGMonitor: fix division by zero on stats dump
1230 Signed-off-by: Joao Eduardo Luis <joao@redhat.com>
1231 (cherry picked from commit 50547dc3c00b7556e26b9a44ec68640c5c3a2384)
1233 commit 640986fdef3543c7469e43853900b63d7b2f6f36
1234 Author: David Zafman <dzafman@redhat.com>
1235 Date: Fri Mar 13 19:16:47 2015 -0700
1237 doc: Fix ceph command manpage to match ceph -h (giant)
1241 Signed-off-by: David Zafman <dzafman@redhat.com>
1243 commit ec4c8997a67397e3ba5f335e444586f6f1a7864c
1244 Author: David Zafman <dzafman@redhat.com>
1245 Date: Fri Mar 13 16:50:13 2015 -0700
1247 doc: Minor fixes to ceph command manpage
1251 Signed-off-by: David Zafman <dzafman@redhat.com>
1252 (cherry picked from commit 7e85722fd4c89715fc2ed79697c82d65d7ebf287)
1254 commit 15e1c6fb3a0e2e191025f96df33feaf115c1411d
1255 Author: David Zafman <dzafman@redhat.com>
1256 Date: Thu Mar 12 11:39:52 2015 -0700
1258 doc: Fix ceph command manpage to match ceph -h (firefly)
1260 Improve synopsis section
1263 Signed-off-by: David Zafman <dzafman@redhat.com>
1264 (cherry picked from commit 9ac488c1eb0e30511079ba05aaf11c79615b3940)
1268 man/ceph.8 (took incoming version)
1270 commit 17d03e8e8b055ed50b1abb679a0dd1edf4ec3f30
1271 Author: Nilamdyuti Goswami <ngoswami@redhat.com>
1272 Date: Thu Dec 18 17:11:22 2014 +0530
1274 doc: Changes format style in ceph to improve readability as html.
1276 Signed-off-by: Nilamdyuti Goswami <ngoswami@redhat.com>
1277 (cherry picked from commit 8b796173063ac9af8c21364521fc5ee23d901196)
1279 commit 61d600687ae6887a2edea0e79d582b1353558a83
1280 Author: Yan, Zheng <zyan@redhat.com>
1281 Date: Tue Mar 10 19:55:57 2015 +0800
1283 mds: fix assertion caused by system clock backwards
1286 Signed-off-by: Yan, Zheng <zyan@redhat.com>
1288 commit f1c9c71fd6d566687db05e1d8f15b04a9f793005
1289 Author: Nilamdyuti Goswami <ngoswami@redhat.com>
1290 Date: Sat Dec 13 02:24:41 2014 +0530
1292 doc: Adds man page for ceph.
1294 Signed-off-by: Nilamdyuti Goswami <ngoswami@redhat.com>
1295 (cherry picked from commit 76da87a64ca6b3cc0ceeaf63e19a9f440d6f4161)
1297 commit 8a5a3c7a20517c7482dc9f5f2783a3da54bd4f23
1298 Author: Dongmao Zhang <deanraccoon@gmail.com>
1299 Date: Fri Nov 14 18:48:58 2014 +0800
1301 Fix libstriprados::stat, use strtoll insdead of strtol
1303 The return value(long int) of strict_strtol is too small for unstriped
1306 Signed-off-by: Dongmao Zhang <deanraccoon@gmail.com>
1307 (cherry picked from commit fe6679dca479fc24806d7e57ab0108a516cd6d55)
1309 commit 5d423a5188c62650eaa39077eb99a84085f5f3e2
1310 Author: Dongmao Zhang <deanraccoon@gmail.com>
1311 Date: Wed Dec 10 18:55:28 2014 +0800
1313 Fix libstriprados::remove, use strtoll insdead of strtol
1315 Signed-off-by: Dongmao Zhang <deanraccoon@gmail.com>
1316 (cherry picked from commit 78a15ee4c61fdadccb1921e861748400cc651862)
1318 commit 938e03630e075af03780da139ae879b5b0377734
1319 Author: Samuel Just <sjust@redhat.com>
1320 Date: Mon Feb 2 13:57:00 2015 -0800
1322 Objecter::_op_submit_with_budget: add timeout before call
1324 Objecter::_send_op depends on the ontimeout field being filled in
1325 to avoid 10340 and 9582.
1328 Signed-off-by: Samuel Just <sjust@redhat.com>
1329 (cherry picked from commit cfcfafcb0f33994dbda1efe478ef3ab822ff50d4)
1331 commit 984df3b865f295ecf77b041a69d1d59384d80671
1332 Author: Kefu Chai <kchai@redhat.com>
1333 Date: Tue Feb 10 16:29:45 2015 +0800
1335 osd: fix PG leak in SnapTrimWQ._clear()
1338 Signed-off-by: Kefu Chai <kchai@redhat.com>
1339 (cherry picked from commit 01e154d592d6cdbf3f859cf1b4357e803536a6b4)
1341 commit c5c6fcd6ad7a0317301970bc25419727ee3192fa
1342 Author: David Zafman <dzafman@redhat.com>
1343 Date: Wed Feb 18 16:21:12 2015 -0800
1345 osd: Fix FileJournal wrap to get header out first
1347 Correct and restore assert that was removed
1349 Cause by f46b1b473fce0322a672b16c7739e569a45054b6
1351 Backport: dumpling, firefly, giant
1353 Signed-off-by: David Zafman <dzafman@redhat.com>
1354 (cherry picked from commit 970bb4901f93575709421b5b25c3eff213de61b8)
1356 commit 37f196e80906d6b855ca0a7ce7344d1f8f50ae7f
1357 Author: Sebastien Ponce <Sebastien.Ponce@cern.ch>
1358 Date: Thu Feb 5 10:46:37 2015 +0100
1360 libradosstriper: fixed write_full when ENOENT
1363 write_full was returning ENOENT when the file did not exists, while it should just have created it without complaining.
1365 Signed-off-by: Sebastien Ponce <sebastien.ponce@cern.ch>
1366 (cherry picked from commit 6a91f2bb82035b3c8021a7dc7b23548ce3d61eb6)
1368 commit b318e2ffc3faedf6b02908429eb4ce79147d3471 (refs/remotes/gh/wip-10643)
1369 Author: Joao Eduardo Luis <joao@redhat.com>
1370 Date: Fri Oct 17 18:01:05 2014 +0100
1372 mon: MDSMonitor: don't return -EINVAL if function is bool
1374 Returning -EINVAL on a function that expects bool and the error code to
1375 be in a variable 'r' can only achieve one thing: if this path is ever
1376 touched, instead of returning an error as it was supposed to, we're
1377 returning 'true' with 'r = 0' and, for no apparent reason, the user will
1378 think everything went smoothly but with no new fs created.
1380 Signed-off-by: Joao Eduardo Luis <joao@redhat.com>
1381 (cherry picked from commit 91b2a2b69b198567d42f45f75b7f7ea559f89150)
1383 commit a1a7d856f5409280b6345afbbd961bcd8ee8d046
1384 Author: Joao Eduardo Luis <joao@redhat.com>
1385 Date: Fri Oct 17 17:43:33 2014 +0100
1387 mon: MDSMonitor: check all conditions are met *before* osdmon proposal
1389 We should not allow ourselves to request the osdmon to propose before we
1390 know for sure that we meet the required conditions to go through with
1391 our own state change. Even if we still can't guarantee that our
1392 proposal is going to be committed, we shouldn't just change the osdmon's
1393 state just because we can. This way, at least, we make sure that our
1394 checks hold up before doing anything with side-effects.
1398 Signed-off-by: Joao Eduardo Luis <joao@redhat.com>
1399 (cherry picked from commit 1d1ae413e9f0aafa314af903bcf6b815f6558082)
1401 commit 0df7cf1e58bc4d77c69354f2a7160e3d9e6d1fc1
1402 Author: Joao Eduardo Luis <joao@redhat.com>
1403 Date: Fri Oct 17 17:37:03 2014 +0100
1405 mon: MDSMonitor: return if fs exists on 'fs new'
1407 We were just setting return code to -EINVAL, while allowing the logic to
1408 continue regardless. If we are to return error, then we should abort
1409 the operation as well and let the user know it went wrong instead of
1410 continuing as if nothing had happened.
1412 Signed-off-by: Joao Eduardo Luis <joao@redhat.com>
1413 (cherry picked from commit 07b7f101057a80d1356bd20040180e94cd20c211)