]> git.proxmox.com Git - ceph.git/blob - ceph/doc/changelog/v0.80.9.txt
add subtree-ish sources for 12.0.3
[ceph.git] / ceph / doc / changelog / v0.80.9.txt
1 commit b5a67f0e1d15385bc0d60a6da6e7fc810bde6047 (tag: refs/tags/v0.80.9, refs/remotes/gh/firefly)
2 Author: Jenkins <jenkins@inktank.com>
3 Date: Mon Mar 9 10:42:08 2015 -0700
4
5 0.80.9
6
7 commit 37901afd1556257151c029395caa1143e84860f2
8 Merge: 00e5947 3dac68a
9 Author: Loic Dachary <ldachary@redhat.com>
10 Date: Tue Mar 3 15:58:53 2015 +0100
11
12 Merge pull request #3852 from dachary/wip-10965-rgw-firefly
13
14 rgw: enable ipv6 in civetweb
15
16 Reviewed-by: Loic Dachary <ldachary@redhat.com>
17
18 commit 00e5947a28f7810a28329d2f4901aed09a289421
19 Merge: 293222a b13f483
20 Author: Loic Dachary <ldachary@redhat.com>
21 Date: Tue Mar 3 15:58:39 2015 +0100
22
23 Merge pull request #3853 from dachary/wip-10907-rgw-firefly
24
25 rgw: pass civetweb configurables to civetweb
26
27 Reviewed-by: Loic Dachary <ldachary@redhat.com>
28
29 commit 293222a284c072a13950831205c106fec7a400df
30 Merge: 6512b06 d57b38f
31 Author: Loic Dachary <ldachary@redhat.com>
32 Date: Tue Mar 3 15:58:25 2015 +0100
33
34 Merge pull request #3851 from dachary/wip-10978-rgw-firefly
35
36 rgw: don't overwrite bucket / object owner when setting acls
37
38 Reviewed-by: Loic Dachary <ldachary@redhat.com>
39
40 commit d57b38f85b683dfb365c3cb98362d486594f9eb3
41 Author: Yehuda Sadeh <yehuda@redhat.com>
42 Date: Fri Feb 27 15:32:50 2015 -0800
43
44 rgw: don't overwrite bucket / object owner when setting acls
45
46 Fixes: #10978
47 Backport: hammer, firefly
48
49 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
50 (cherry picked from commit eb13f2d4b60c031f16139f7cc4237c012644dd78)
51
52 Conflicts:
53 src/rgw/rgw_op.cc : trivial s/.empty()/== NULL/
54
55 commit b13f483e51b37a768c7f4313b6933bf648950c7d
56 Author: Yehuda Sadeh <yehuda@redhat.com>
57 Date: Tue Feb 17 15:05:40 2015 -0800
58
59 rgw: pass civetweb configurables to civetweb
60
61 Fixes: #10907
62 Backport: firefly
63
64 Pass any configurables in the rgw frontends config line to civetweb.
65
66 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
67 (cherry picked from commit 986d7554426764a149621ba733c5c075b94e0431)
68
69 commit 3dac68a17a909b212a36b0a3c0ae2c47d323deee
70 Author: Yehuda Sadeh <yehuda@redhat.com>
71 Date: Fri Feb 27 08:14:27 2015 -0800
72
73 civetweb: update submodule
74
75 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
76 (cherry picked from commit f28fe744285117a1715eac0d08911cdb37285103)
77
78 commit 14aca3af18119a76c1cdfa6d71e6085d360e45e2
79 Author: Yehuda Sadeh <yehuda@redhat.com>
80 Date: Fri Feb 27 08:14:41 2015 -0800
81
82 rgw: update makefile to enable civetweb config
83
84 Fixes: #10965
85 Backport: hammer, firefly
86
87 Civetweb compilation now includes conf header to enable ipv6.
88
89 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
90 (cherry picked from commit caa90225bad9fe5e9c275e6189b3396b4d396e3f)
91
92 Conflicts:
93 src/rgw/Makefile.am
94 radosgw_CFLAGS = -Icivetweb/include
95 was not yet changed into
96 radosgw_CFLAGS = -I$(srcdir)/civetweb/include
97 in firefly
98
99 commit 6512b06fc8a80e3befbe15b543f8850991f74e8a
100 Merge: 7b748c6 555cc42
101 Author: Loic Dachary <ldachary@redhat.com>
102 Date: Sat Feb 28 16:47:55 2015 +0100
103
104 Merge pull request #3820 from tchaikov/firefly-pg-leak-10421
105
106 osd: fix PG leak in SnapTrimWQ._clear()
107
108 Reviewed-by: Loic Dachary <ldachary@redhat.com>
109
110 commit 7b748c62764311572342593820ce3f28f74fe9ca
111 Merge: d971c95 da95149
112 Author: Loic Dachary <ldachary@redhat.com>
113 Date: Sat Feb 28 16:46:42 2015 +0100
114
115 Merge pull request #3772 from ceph/wip-10883-firefly
116
117 osd: Fix FileJournal wrap to get header out first
118
119 Reviewed-by: Loic Dachary <ldachary@redhat.com>
120
121 commit d971c95c3d8b48494d05b68f40c4c9a4cc6f87a8
122 Merge: 0820041 e539971
123 Author: Loic Dachary <ldachary@redhat.com>
124 Date: Sat Feb 28 16:45:48 2015 +0100
125
126 Merge pull request #3521 from dzafman/wip-10676
127
128 Backport doc fixes that appear to apply to firefly
129
130 Reviewed-by: Loic Dachary <ldachary@redhat.com>
131
132 commit 555cc42fc826fd801f0d45187429079d4072d129
133 Author: Kefu Chai <kchai@redhat.com>
134 Date: Tue Feb 10 16:29:45 2015 +0800
135
136 osd: fix PG leak in SnapTrimWQ._clear()
137
138 Fixes: #10421
139 Signed-off-by: Kefu Chai <kchai@redhat.com>
140 (cherry picked from commit 01e154d592d6cdbf3f859cf1b4357e803536a6b4)
141
142 commit 0820041e6515c82c41b81b9e6825e2dd5fcb8165
143 Merge: 6565774 702dbc0
144 Author: Gregory Farnum <greg@gregs42.com>
145 Date: Thu Feb 26 15:59:24 2015 -0800
146
147 Merge pull request #3730 from ceph/wip-firefly-flock
148
149 backport ceph-fuse file locking patches to Firefly
150
151 Reviewed-by: Yan, Zheng <zyan@redhat.com>
152
153 commit 6565774d0356efc6225ad561bb13e7cf11da7b1e
154 Merge: 7ab02ee 08c2fda
155 Author: Josh Durgin <jdurgin@redhat.com>
156 Date: Thu Feb 26 14:18:10 2015 -0800
157
158 Merge pull request #3406 from ceph/wip-10299-firefly
159
160 librbd: complete all pending aio ops prior to closing image
161
162 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
163
164 commit 08c2fda12cf46937a09a59bb032379c3c5321292
165 Author: Jason Dillaman <dillaman@redhat.com>
166 Date: Mon Dec 15 10:53:53 2014 -0500
167
168 librbd: complete all pending aio ops prior to closing image
169
170 It was possible for an image to be closed while aio operations
171 were still outstanding. Now all aio operations are tracked and
172 completed before the image is closed.
173
174 Fixes: #10299
175 Backport: giant, firefly, dumpling
176 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
177
178 commit 7ab02ee5afb3d017b94d58c3dfc7731f7a3866d9
179 Merge: 0c0a552 c23e42e
180 Author: Josh Durgin <jdurgin@redhat.com>
181 Date: Thu Feb 26 14:12:29 2015 -0800
182
183 Merge pull request #3404 from ceph/wip-10270-firefly
184
185 librbd: gracefully handle deleted/renamed pools
186
187 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
188
189 commit 0c0a5520b1b883bcdd5b865b217ba61e471e3ca2
190 Merge: 9ef7743 836ab86
191 Author: Josh Durgin <jdurgin@redhat.com>
192 Date: Thu Feb 26 13:31:37 2015 -0800
193
194 Merge pull request #3410 from ceph/wip-9854-firefly
195
196 osdc: Constrain max number of in-flight read requests
197
198 Reviewed-by: Josh Durgin <jdurgin@redhat.com>
199
200 commit da951497b7e5ec227aa6a5e459b0d04d8b88ef13 (refs/remotes/gh/wip-10883-firefly)
201 Author: David Zafman <dzafman@redhat.com>
202 Date: Wed Feb 18 16:21:12 2015 -0800
203
204 osd: Fix FileJournal wrap to get header out first
205
206 Correct and restore assert that was removed
207
208 Cause by f46b1b473fce0322a672b16c7739e569a45054b6
209 Fixes: #10883
210 Backport: dumpling, firefly, giant
211
212 Signed-off-by: David Zafman <dzafman@redhat.com>
213 (cherry picked from commit 970bb4901f93575709421b5b25c3eff213de61b8)
214
215 commit 702dbc0a247c149d53b52d1929f9880bc99d0522
216 Author: Yan, Zheng <zyan@redhat.com>
217 Date: Wed Oct 15 12:00:58 2014 +0800
218
219 qa/workunits/fs/misc: Add a workunit for file lock interruption
220
221 Signed-off-by: Yan, Zheng <zyan@redhat.com>
222 (cherry picked from commit ac92c455a9aa19e4288acdf0c9a746e03a640efb)
223
224 commit d2523b82c5c7b29293d70a66ba95493a1564a840
225 Author: Yan, Zheng <zyan@redhat.com>
226 Date: Wed Oct 15 12:03:46 2014 +0800
227
228 mds: fix neighbor lock check
229
230 Signed-off-by: Yan, Zheng <zyan@redhat.com>
231 (cherry picked from commit b0e6e85aa08ea74cd209aad04f3f0bf991761e12)
232
233 commit f9b6b66b05ddadef043d81676728bf40730ea16c
234 Author: Yan, Zheng <zyan@redhat.com>
235 Date: Mon Oct 13 11:34:18 2014 +0800
236
237 client: use finisher to abort MDS request
238
239 When a request is interrupted, libfuse first locks an internal mutex,
240 then calls the interrupt callback. libfuse need to lock the same mutex
241 when unregistering interrupt callback. We unregister interrupt callback
242 while client_lock is locked, so we can't acquiring the client_lock in
243 the interrupt callback.
244
245 Signed-off-by: Yan, Zheng <zyan@redhat.com>
246 (cherry picked from commit 09699454e729592d426aeff5b578697e850af12e)
247
248 Conflicts:
249 src/client/Client.cc
250 src/client/Client.h
251
252 Signed-off-by: Greg Farnum <gfarnum@redhat.com>
253
254 commit ea355e9ca5203b77e6f74ca4a3e39ce23cc86f67
255 Author: Yan, Zheng <zyan@redhat.com>
256 Date: Thu Oct 9 13:16:18 2014 +0800
257
258 client: use atomic variable to track reference of MetaRequeset
259
260 this allow us to increase reference count of MetaRequest while not holding
261 the client_lock
262
263 Signed-off-by: Yan, Zheng <zyan@redhat.com>
264 (cherry picked from commit e464a7765230c504b3e3b88bcb1106f67b7c3eb9)
265
266 commit ccbdf514717cb0e48d67b57197d6e538faeea415
267 Author: Yan, Zheng <zyan@redhat.com>
268 Date: Mon Oct 13 10:44:46 2014 +0800
269
270 client: allow interrupting blocked file lock operation
271
272 This commit introduce two new types of setfilelock request. Unlike
273 setfilelock (UNLOCK) request, these two new types of setfilelock request
274 do not drop locks that have alread been acquired, they only interrupt
275 blocked setfilelock request.
276
277 Signed-off-by: Yan, Zheng <zyan@redhat.com>
278 (cherry picked from commit 4134c149d3759dd6a3aaa1a353b77bbfe8e9491b)
279
280 commit 875e2fcb060554941d94714ad48ebbc0cbbf8077
281 Author: Yan, Zheng <zyan@redhat.com>
282 Date: Thu Oct 9 09:42:08 2014 +0800
283
284 client: register callback for fuse interrupt
285
286 libfuse allows program to reigster a callback for interrupt. When a file
287 system operation is interrupted, the fuse kernel driver sends interupt
288 request to libfuse. libfuse calls the interrupt callback when receiving
289 interrupt request.
290
291 Signed-off-by: Yan, Zheng <zyan@redhat.com>
292 (cherry picked from commit 289e8b4a7efa1ae6427115af9bbe541c9e1f0e90)
293
294 commit c96aabbd54b24a0273af21a450cd6f517fe4ada0
295 Author: Yan, Zheng <zyan@redhat.com>
296 Date: Sat Oct 4 09:14:44 2014 +0800
297
298 client: add helper function that updates lock state
299
300 Signed-off-by: Yan, Zheng <zyan@redhat.com>
301 (cherry picked from commit 6a2303a6b6d97f2a6d1422e42d3d88991857618f)
302
303 Conflicts:
304 src/client/Client.h
305
306 Signed-off-by: Greg Farnum <gfarnum@redhat.com>
307
308 commit ebbd3ca66722cdc61b56d243baf8e63f7b8c1c1b
309 Author: Yan, Zheng <zyan@redhat.com>
310 Date: Thu Oct 2 20:21:36 2014 +0800
311
312 fuse: enable fuse_multithreaded by default
313
314 GETFILELOCK MDS request may block for a long time, so we need to
315 use multithread event loop,
316
317 Signed-off-by: Yan, Zheng <zyan@redhat.com>
318 (cherry picked from commit b17b43a8660ed0db29fbecf44798265e47712f85)
319
320 commit 924e6f2b6b03456a8b18d8b158b8f325051f0519
321 Author: Yan, Zheng <zyan@redhat.com>
322 Date: Thu Oct 2 19:07:41 2014 +0800
323
324 client: posix file lock support
325
326 Signed-off-by: Yan, Zheng <zyan@redhat.com>
327 (cherry picked from commit a1b2c8ff955b30807ac53ce6bdc97cf61a7262ca)
328
329 Conflicts:
330 src/client/Client.cc
331 src/client/Client.h
332 src/client/Inode.h
333
334 Signed-off-by: Greg Farnum <gfarnum@redhat.com>
335
336 commit 82994946bdcb95867a61e0acf443e30b13925a34
337 Author: Yan, Zheng <zyan@redhat.com>
338 Date: Thu Oct 2 18:02:50 2014 +0800
339
340 common: link mds/flock.o to libcommon
341
342 later commit will use this code to track file locks held by cephfs
343 client.
344
345 Signed-off-by: Yan, Zheng <zyan@redhat.com>
346 (cherry picked from commit e075c27c3554380c59dce0cc17ef0944eb415025)
347
348 Conflicts:
349 src/mds/Makefile.am
350
351 Signed-off-by: Greg Farnum <gfarnum@redhat.com>
352
353 commit 9ef77430f3d46789b0ba1a2afa42729627734500
354 Merge: 1205867 b668566
355 Author: Loic Dachary <ldachary@redhat.com>
356 Date: Fri Feb 13 18:34:11 2015 +0100
357
358 Merge pull request #3684 from ceph/wip-crush-straw-firefly
359
360 osd: backport straw bucket fixes and all other recent crush goodness to firefly
361
362 Reviewed-by: Loic Dachary <ldachary@redhat.com>
363
364 commit b668566fd8148414b8074f096b85b22c42ed3af9 (refs/remotes/gh/wip-crush-straw-firefly)
365 Author: Sage Weil <sage@redhat.com>
366 Date: Fri Jan 16 09:02:28 2015 -0800
367
368 crush/builder: fix warnings
369
370 crush/builder.c: In function 'crush_remove_list_bucket_item':
371 crush/builder.c:977:13: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
372 if (weight < bucket->h.weight)
373 ^
374 crush/builder.c: In function 'crush_remove_tree_bucket_item':
375 crush/builder.c:1031:14: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
376 if (weight < bucket->h.weight)
377 ^
378
379 Signed-off-by: Sage Weil <sage@redhat.com>
380 (cherry picked from commit 14eb1a73c71d81b7f193fce27c59cb3babf3e74a)
381
382 commit 247afa4efe7d11b5ef06e094680b50ea30d7d5e3
383 Author: Loic Dachary <loic-201408@dachary.org>
384 Date: Wed Oct 15 17:02:58 2014 -0700
385
386 crush: improve constness of CrushWrapper methods
387
388 A number of CrushWrapper get methods or predicates were not const
389 because they need to maintain transparently the rmaps. Make the rmaps
390 mutable and update the constness of the methods to match what the caller
391 would expect.
392
393 Signed-off-by: Loic Dachary <loic-201408@dachary.org>
394 (cherry picked from commit 236895eea65f8706baa5fdef96fb00ad5b82218c)
395
396 commit c59279a25b6c53e0ab3988b0f00ae3cce94f33d7
397 Author: Xiaoxi Chen <xiaoxi.chen@intel.com>
398 Date: Fri Sep 5 10:56:36 2014 +0800
399
400 Change CrushWrapper::crush to private
401
402 Currently in CrushWrapper, the member "struct crush_map *crush" is a public member,
403 so people can break the encapsulation and manipulate directly to the crush structure.
404
405 This is not a good practice for encapsulation and will lead to inconsistent if code
406 mix use the CrushWrapper API and crush C API.A simple example could be:
407 1.some code use crush_add_rule(C-API) to add a rule, which will not set the have_rmap flag to false in CrushWrapper
408 2.another code using CrushWrapper trying to look up the newly added rule by name will get a -ENOENT.
409
410 This patch move CrushWrapper::crush to private, together with three reverse map(type_rmap, name_rmap, rule_name_rmap)
411 and also change codes accessing the CrushWrapper::crush to make it compile.
412
413 Signed-off-by: Xiaoxi Chen <xiaoxi.chen@intel.com>
414 (cherry picked from commit d734600f9251b52f525faa35441e2b5dd660161b)
415
416 commit 9fc2fd50aec94233528870aaa371347238a6ea75
417 Author: Sage Weil <sage@inktank.com>
418 Date: Fri Jul 11 06:58:57 2014 -0700
419
420 crush: include CRUSH_V3, v2/v3 rules checks in dump_tunables()
421
422 Backport: firefly
423 Signed-off-by: Sage Weil <sage@inktank.com>
424 (cherry picked from commit cf94cf3531a349bbd1fc6ee56c3fc260110a252a)
425
426 commit f6009a614a3689cc9a96f55b35e70362be4ec64b
427 Author: Sage Weil <sage@redhat.com>
428 Date: Fri Dec 5 15:58:03 2014 -0800
429
430 mon: 'osd crush reweight-all'
431
432 This corresponds to the crushtool --reweight command.
433
434 Signed-off-by: Sage Weil <sage@redhat.com>
435 (cherry picked from commit 89b2feea8d53b9dc15ab5ae7f5920ad19c8bba18)
436
437 commit d5d7495432dd023ed8c9ba2ac08222f06afee48f
438 Author: Sage Weil <sage@redhat.com>
439 Date: Fri Dec 5 15:55:24 2014 -0800
440
441 crush: set straw_calc_version=1 for default+optimal; do not touch for presets
442
443 When using the presets for compatibility (i.e., based on version), do not
444 touch the straw behavior, as it does not affect mapping or compatibility.
445 However, make a point of setting it by default and for optimal.
446
447 For most users, this means that they will not see any change unless they
448 explicitly enable the new behavior, or switch to default or optimal
449 tunables. The idea is that if they touched it, they shouldn't be
450 too surprised by the subsequent data movement.
451
452 Signed-off-by: Sage Weil <sage@redhat.com>
453 (cherry picked from commit dd7b58f3b1aa1febfc6dc227937df93ee6e284eb)
454
455 (Note: differs from original in that the hammer tunable profile is not
456 present)
457
458 commit 5b7b7599123a9d7a837ded52946bd553b09b1ac7
459 Author: Sage Weil <sage@redhat.com>
460 Date: Wed Dec 3 22:30:00 2014 -0800
461
462 crush/builder: a note about the original crush_calc_straw()
463
464 Signed-off-by: Sage Weil <sage@redhat.com>
465 (cherry picked from commit adf5c6de0aca24a53d3c7b4e7eeb0a5dce9db0f1)
466
467 commit 3ab835b059fd74a525cc2a8ebe8b6a1453e0cc87
468 Author: Sage Weil <sage@redhat.com>
469 Date: Tue Dec 2 16:43:16 2014 -0800
470
471 mon: add 'osd crush {get,set}-tunable <name> [value]' commands
472
473 For now, just add the straw_calc_version tunable.
474
475 Signed-off-by: Sage Weil <sage@redhat.com>
476 (cherry picked from commit 9000068ae45a8b89315c152b7d5509ac873f2957)
477
478 Conflicts:
479 src/mon/OSDMonitor.cc
480
481 commit 884414c5788bac9a269f01b26cbc0c55850c34f6
482 Author: Sage Weil <sage@redhat.com>
483 Date: Tue Dec 2 16:33:11 2014 -0800
484
485 crush: fix crush_calc_straw() scalers when there are duplicate weights
486
487 The straw bucket was originally tested with uniform weights and with a
488 few more complicated patterns, like a stair step (1,2,3,4,5,6,7,8,9). And
489 it worked!
490
491 However, it does not behave with a pattern like
492 1, 2, 2, 3, 3, 4, 4
493
494 Strangely, it does behave with
495 1, 1, 2, 2, 3, 3, 4, 4
496
497 and more usefully it does behave with
498 1, 2, 2.001, 3, 3.001, 4, 4.001
499
500 That is, the logic that explicitly copes with weights that are duplicates
501 is broken.
502
503 The fix is to simply remove the special handling for duplicate weights --
504 it isn't necessary and doesn't work correctly anyway.
505
506 Add a test that compares the mapping result of [1, 2, 2, 3, 3, ...] with
507 [1, 2, 2.001, 3, 3.001, ...] and verifies that the difference is small.
508 With the fix, we get .00012, whereas the original implementation gets
509 .015.
510
511 Note that this changes the straw bucket scalar *precalculated* values that
512 are encoded with the map, and only when the admin opts into the new behavior.
513
514 Backport: giant, firefly
515 Signed-off-by: Sage Weil <sage@redhat.com>
516 (cherry picked from commit 43d5c7caa7ce478477bde1bbd4f0649b5159cdcf)
517
518 commit bf677093302f475a23de541471ddde6a7c7153b6
519 Author: Sage Weil <sage@redhat.com>
520 Date: Tue Dec 2 14:50:21 2014 -0800
521
522 crush: fix distortion of straw scalers by 0-weight items
523
524 The presence of a 0-weight item in a straw bucket should have no effect
525 on the placement of other items. Add a test validating that and fix
526 crush_calc_straw() to fix the distortion.
527
528 Note that this effects the *precalculation* of the straw bucket inputs and
529 does not effect the actually mapping process given a compiled or encoded
530 CRUSH map, and only when straw_calc_version == 1 (i.e., the admin opted in
531 to the new behavior).
532
533 Backport: giant, firefly
534 Signed-off-by: Sage Weil <sage@redhat.com>
535 (cherry picked from commit 85498bc8f62ca56506b33f3c5ec4fc4b111ed73d)
536
537 commit e02574ef5d3b66e73424a16460366de2d36eded3
538 Author: Sage Weil <sage@redhat.com>
539 Date: Tue Dec 2 14:49:42 2014 -0800
540
541 crush/builder: break out new version 1 of crush_calc_straw
542
543 No change, yet.
544
545 Signed-off-by: Sage Weil <sage@redhat.com>
546 (cherry picked from commit 7c1203635f2b5911f281ce3a441905df6e9bd103)
547
548 commit c7312a47865c758e67852999803d8aa90ff809c1
549 Author: Sage Weil <sage@redhat.com>
550 Date: Tue Dec 2 14:45:04 2014 -0800
551
552 crush: pass crush_map * to various builder methods
553
554 In particular, we will need it for crush_calc_straw().
555
556 Signed-off-by: Sage Weil <sage@redhat.com>
557 (cherry picked from commit f35a3d88cb944c292e966f679ac7a8d7a1cd3093)
558
559 Conflicts:
560 src/crush/CrushWrapper.cc
561
562 commit 5137cc656548d942d0f21ba3ef28a5b7d4d21831
563 Author: Rongze Zhu <zrzhit@gmail.com>
564 Date: Fri Oct 10 19:18:00 2014 +0800
565
566 crush: fix incorrect use of adjust_item_weight method
567
568 adjust_item_weight method will adjust all buckets which the item
569 inside. If the osd.0 in host=fake01 and host=fake02, we execute
570 "ceph osd crush osd.0 10 host=fake01", it not only will adjust fake01's
571 weight, but also will adjust fake02's weight.
572
573 the patch add adjust_item_weightf_in_loc method and fix remove_item,
574 _remove_item_under, update_item, insert_item, detach_bucket methods.
575
576 Signed-off-by: Rongze Zhu <zrzhit@gmail.com>
577 (cherry picked from commit 9850227d2f0ca2f692a154de2c14a0a08e751f08)
578
579 Conflicts:
580 src/crush/CrushWrapper.cc
581
582 commit 80ec22014deb8536b9c34e8d57b286052898d05a
583 Author: Sage Weil <sage@redhat.com>
584 Date: Thu Nov 13 10:59:22 2014 -0800
585
586 crush/CrushWrapper: fix detach_bucket
587
588 In commit 9850227d2f0ca2f692a154de2c14a0a08e751f08 we changed the call that
589 changed the weight of all instances of item to one that explicitly
590 changes it in the parent bucket, but parent_id may not be valid at the
591 call site. Move this into the conditional block to fix.
592
593 Fixes: #10095
594 Signed-off-by: Sage Weil <sage@redhat.com>
595 (cherry picked from commit 6f8b96a9fe4793906c74a571109a457aca7ca220)
596
597 commit 5bc554ac5a7a4cd86b873bfc4f4313c91287b52c
598 Author: Sage Weil <sage@redhat.com>
599 Date: Tue Dec 2 14:10:49 2014 -0800
600
601 crush: default to straw_calc_version 1
602
603 Stick with bobtail tunables where it affects compatibility. Use v1 of
604 straw_calc, though, since that does not, and we want the best for new
605 clusters.
606
607 Signed-off-by: Sage Weil <sage@redhat.com>
608 (cherry picked from commit 9565621d6007302fdd68ba27b6aef22e487f0985)
609
610 commit cfc718a5a1fec937cf00e6c2b55f66d4390088bb
611 Author: Sage Weil <sage@redhat.com>
612 Date: Fri Feb 13 08:30:35 2015 -0800
613
614 crush: add straw_calc_version tunable
615
616 It doesn't do anything, yet.
617
618 Signed-off-by: Sage Weil <sage@redhat.com>
619 (cherry picked from commit 50e2ceefee6a5dfbecbe54890139b1fa80a313c8)
620
621 (Modified from original to not create the 'hammer' tunable profile, which
622 we will not backport in its entirety.)
623
624 commit 1205867d3afe6d457483f9f51aaee79cca74e040
625 Author: Josh Durgin <jdurgin@redhat.com>
626 Date: Mon Feb 2 16:43:35 2015 +0100
627
628 qa: use correct binary path on rpm-based systems
629
630 Fixes: #10715
631 Signed-off-by: Josh Durgin <jdurgin@redhat.com>
632 (cherry picked from commit 05ce2aa1bf030ea225300b48e7914577a412b38c)
633
634 commit da4146a5e03503083e6bc5c12984c06f41a3b4be
635 Author: Greg Farnum <gfarnum@redhat.com>
636 Date: Thu Feb 5 21:12:17 2015 -0800
637
638 fsync-tester: print info about PATH and locations of lsof lookup
639
640 We're seeing the lsof invocation fail (as not found) in testing and nobody can
641 identify why. Since attempting to reproduce the issue has not worked, this
642 patch will gather data from a genuinely in-vitro location.
643
644 Signed-off-by: Greg Farnum <gfarnum@redhat.com>
645 (cherry picked from commit a85051483874ff5b8b0fb50426a3577040457596)
646
647 commit 61b4f09848796faeacf7eb78dd6dc06513beb737
648 Author: Sage Weil <sage@redhat.com>
649 Date: Tue Dec 2 14:04:34 2014 -0800
650
651 crush/CrushWrapper: dump chooseleaf_vary_r with other tunables
652
653 Signed-off-by: Sage Weil <sage@redhat.com>
654 (cherry picked from commit c133a83fe7b77e2a7e7c711cb8ab943c59ff0885)
655
656 commit e4939ed5535cc4678cf1d1ae80906290448e6590
657 Author: Sage Weil <sage@redhat.com>
658 Date: Tue Dec 2 10:12:25 2014 -0800
659
660 crush/CrushTester: add new --show-mappings option
661
662 This makes --show-utilization and --show-statistics usable.
663
664 Signed-off-by: Sage Weil <sage@redhat.com>
665 (cherry picked from commit 3877f00877f53e9f86630c2d89e81ab9cff729ec)
666
667 commit 8c48ebbf13e6bdb689b4b8ae58ac811653ad2acc
668 Author: Sage Weil <sage@redhat.com>
669 Date: Tue Dec 2 10:11:57 2014 -0800
670
671 crushtool/CrushTester: output utilization even with 1 batch
672
673 Signed-off-by: Sage Weil <sage@redhat.com>
674 (cherry picked from commit 294b06c6424f4cb69394976add826d9725073b50)
675
676 commit e9faab9f296af9cf26b9475afd3562c3f3b9236b
677 Author: Sage Weil <sage@redhat.com>
678 Date: Tue Dec 2 10:08:18 2014 -0800
679
680 crush: recalculate straw scalers during a reweight
681
682 The crushtool --reweight function triggers a fresh calculation of bucket
683 weights so that they are always the sum of the item weights. In the
684 straw bucket case, the weights were updated but the corresponding straw
685 scalers were not being recalculated. The result is that there was not
686 effect on placement in adjusted buckets until the next time a bucket item's
687 weight was adjusted.
688
689 Backport: giant, firefly
690 Signed-off-by: Sage Weil <sage@redhat.com>
691 (cherry picked from commit 35062937f174a45fb13d9c177eaa1fe4ed5ff4c2)
692
693 commit 4c951e9dc1ca31429119de15755caf3d73f6ffce
694 Author: Sage Weil <sage@redhat.com>
695 Date: Tue Dec 2 08:36:41 2014 -0800
696
697 osdmaptool: --test-map-pgs-dump
698
699 Signed-off-by: Sage Weil <sage@redhat.com>
700 (cherry picked from commit 7a99b489909b368bade36d8cc9722ad33d03a2e8)
701
702 commit 3cb5d08feee4811624dd495f74912824203edf00
703 Author: Sage Weil <sage@redhat.com>
704 Date: Tue Dec 2 06:53:20 2014 -0800
705
706 crush: add dprintk's for crush_calc_straw
707
708 These are compiled out by default.
709
710 Signed-off-by: Sage Weil <sage@redhat.com>
711 (cherry picked from commit 946bebd94b109642f95710802e892c59eb4f0b76)
712
713 commit ca8471d65fb2cf7d2247d823c5436faa273efda5
714 Author: Rongze Zhu <zrzhit@gmail.com>
715 Date: Tue Nov 11 00:13:42 2014 +0800
716
717 crush: fix tree bucket functions
718
719 There are incorrect nodes' weight in tree bucket when construct tree
720 bucket. The tree bucket don't store item id in items array, so the tree
721 bucket will not work correctly. The patch fix above bugs and add a
722 simple test for tree bucket.
723
724 Signed-off-by: Rongze Zhu <zrzhit@gmail.com>
725 (cherry picked from commit 13425488882d360fa740613dfcfd0d098c1b7616)
726
727 commit af502f25b04cd0758b753941ecf0b1b59d33ea9e
728 Author: Sage Weil <sage@redhat.com>
729 Date: Tue Nov 11 11:16:58 2014 -0800
730
731 crush/builder: replace printf with an empty dprintk macro
732
733 This mirrors mapper.c.
734
735 Signed-off-by: Sage Weil <sage@redhat.com>
736 (cherry picked from commit e444b221733360cdea4f4bbdbbbfbf2b6ee30ff5)
737
738 commit 26966c4aa9cd79cb53db052553a5fc5653f2591b
739 Author: Greg Farnum <gfarnum@redhat.com>
740 Date: Thu Nov 6 17:48:01 2014 -0800
741
742 qa: use sudo even more when rsyncing /usr
743
744 Signed-off-by: Greg Farnum <gfarnum@redhat.com>
745 (cherry picked from commit 3aa7797741f9cff06053a2f31550fe6929039692)
746
747 commit 6af48421c0ba6195e9a1607053c42eeb62b14ccb
748 Author: Greg Farnum <greg@inktank.com>
749 Date: Tue Oct 21 10:55:06 2014 -0700
750
751 qa: use sudo when rsyncing /usr so we can read everything
752
753 Signed-off-by: Greg Farnum <greg@inktank.com>
754 (cherry picked from commit fa07c04231db2d130de54647957ffab4a7a53733)
755
756 commit 861a18b1f7dd5bdcbb85bc3fa15d1719bb4b2813
757 Author: Yehuda Sadeh <yehuda@redhat.com>
758 Date: Thu Jan 15 16:31:22 2015 -0800
759
760 rgw: fix partial GET in swift
761
762 Fixes: #10553
763 backport: firefly, giant
764
765 Don't set the ret code to reflect partial download, just set the
766 response status when needed.
767
768 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
769 (cherry picked from commit 7e1553cedff90fa0fefded65cde87ad068eb5f0c)
770
771 commit e539971e2d528b4de6009ea44565f037acb2be66 (refs/remotes/gh/wip-10676)
772 Author: Nilamdyuti Goswami <ngoswami@redhat.com>
773 Date: Thu Dec 18 17:13:27 2014 +0530
774
775 doc: Adds updated man page for ceph under man/
776
777 Signed-off-by: Nilamdyuti Goswami <ngoswami@redhat.com>
778 (cherry picked from commit 8de9a0f437822c770600c19a9f61977745b7e530)
779
780 commit 15596ffe388147b7984457041a38cbb9f472556c
781 Author: Nilamdyuti Goswami <ngoswami@redhat.com>
782 Date: Sat Dec 13 02:27:45 2014 +0530
783
784 doc: Adds man page for ceph under man/.
785
786 Signed-off-by: Nilamdyuti Goswami <ngoswami@redhat.com>
787 (cherry picked from commit ffd6c7e49686f8f92ddb400ffdec62520708e64b)
788
789 commit 893e5cd1f4fd5ea77d65f0e78cc8cff88eb19c1a
790 Author: Nilamdyuti Goswami <ngoswami@redhat.com>
791 Date: Sat Dec 13 02:24:41 2014 +0530
792
793 doc: Adds man page for ceph.
794
795 Signed-off-by: Nilamdyuti Goswami <ngoswami@redhat.com>
796 (cherry picked from commit 76da87a64ca6b3cc0ceeaf63e19a9f440d6f4161)
797
798 commit 8d29a4a231666830914903b95599d80da7b97def
799 Author: Sage Weil <sage@redhat.com>
800 Date: Mon Dec 15 17:04:32 2014 -0800
801
802 osd: handle no-op write with snapshot case
803
804 If we have a transaction that does something to the object but it !exists
805 both before and after, we will continue through the write path. If the
806 snapdir object already exists, and we try to create it again, we will
807 leak a snapdir obc and lock and later crash on an assert when the obc
808 is destroyed:
809
810 0> 2014-12-06 01:49:51.750163 7f08d6ade700 -1 osd/osd_types.h: In function 'ObjectContext::~ObjectContext()' thread 7f08d6ade700 time 2014-12-06 01:49:51.605411
811 osd/osd_types.h: 2944: FAILED assert(rwstate.empty())
812
813 Fix is to not recreated the snapdir if it already exists.
814
815 Fixes: #10262
816 Signed-off-by: Sage Weil <sage@redhat.com>
817 (cherry picked from commit 02fae9fc54c10b5a932102bac43f32199d4cb612)
818
819 commit 8ba48d10c252b28cde5b4da1286421db12b57cc2
820 Author: Sage Weil <sage@redhat.com>
821 Date: Mon Jan 19 18:28:20 2015 -0800
822
823 ceph_test_rados_api_misc: do not assert rbd feature match
824
825 This test fails on upgrades when we (or the server) have new
826 features. Make it less fragile.
827
828 Fixes: #10576
829 Signed-off-by: Sage Weil <sage@redhat.com>
830 (cherry picked from commit 9147c62989871cea8b3a85b02c53017825efb55b)
831
832 commit 836ab86d89ae15ed5b228ff656bc81c7cc3495aa
833 Author: Jason Dillaman <dillaman@redhat.com>
834 Date: Mon Oct 27 14:47:19 2014 -0400
835
836 osdc: Constrain max number of in-flight read requests
837
838 Constrain the number of in-flight RADOS read requests to the
839 cache size. This reduces the chance of the cache memory
840 ballooning during certain scenarios like copy-up which can
841 invoke many concurrent read requests.
842
843 Fixes: #9854
844 Backport: giant, firefly, dumpling
845 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
846 (cherry picked from commit 068d68850d09dfcaccc5a3ce85c80b2f6d808ea9)
847
848 commit 2964efaf327cf798f892a6722eb4e24f2ffa0fde
849 Author: Jason Dillaman <dillaman@redhat.com>
850 Date: Mon Jan 19 12:46:00 2015 -0500
851
852 Revert "Enforce cache size on read requests"
853
854 This reverts commit b7784dc1baa47560a733fe9dcd2acec51bc93165.
855
856 commit e4c5b153735aba3f84554a50243f779f36568373
857 Author: Jason Dillaman <dillaman@redhat.com>
858 Date: Mon Jan 19 12:45:25 2015 -0500
859
860 Revert "rbd: ObjectCacher reads can hang when reading sparse files"
861
862 This reverts commit 711a7e6f81983ff2091caa0f232af914a04a041c.
863
864 commit c23e42e78ea3ba44706951a728e2ccb74cac7b33
865 Author: Jason Dillaman <dillaman@redhat.com>
866 Date: Mon Jan 19 10:28:56 2015 -0500
867
868 librbd: gracefully handle deleted/renamed pools
869
870 snap_unprotect and list_children both attempt to scan all
871 pools. If a pool is deleted or renamed during the scan,
872 the methods would previously return -ENOENT. Both methods
873 have been modified to more gracefully handle this condition.
874
875 Fixes: #10270
876 Backport: giant, firefly
877 Signed-off-by: Jason Dillaman <dillaman@redhat.com>
878 (cherry picked from commit 436923c68b77c900b7774fbef918c0d6e1614a36)
879
880 commit 24c13d87039d4f61df0bcabdb8862e0e94fe575d
881 Author: Yehuda Sadeh <yehuda@redhat.com>
882 Date: Fri Dec 12 05:24:01 2014 -0800
883
884 rgw: change multipart upload id magic
885
886 Fixes: #10271
887 Backport: firefly, giant
888
889 Some clients can't sign requests correctly with the original magic
890 prefix.
891
892 Reported-by: Georgios Dimitrakakis <giorgis@acmac.uoc.gr>
893 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
894 (cherry picked from commit 5fc7a0be67a03ed63fcc8408f8d71a31a1841076)
895
896 commit 617002d3ff469ef409a83e35d4f4fd6a0b5b1278
897 Author: Yehuda Sadeh <yehuda@redhat.com>
898 Date: Thu Dec 11 09:07:10 2014 -0800
899
900 rgw: url decode http query params correctly
901
902 Fixes: #10271
903 Backport: firefly
904
905 This got broken by the fix for #8702. Since we now only url_decode if
906 we're in query, we need to specify that we're in query when decoding
907 these args.
908
909 Reported-by: Georgios Dimitrakakis <giorgis@acmac.uoc.gr>
910 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
911 (cherry picked from commit 21e07eb6abacb085f81b65acd706b46af29ffc03)
912
913 commit d7ccf71d962cec1571f53c9392f9b58350569062
914 Author: Josh Durgin <jdurgin@redhat.com>
915 Date: Wed Jan 14 15:01:38 2015 -0800
916
917 qa: ignore duplicates in rados ls
918
919 These can happen with split or with state changes due to reordering
920 results within the hash range requested. It's easy enough to filter
921 them out at this stage.
922
923 Backport: giant, firefly
924 Signed-off-by: Josh Durgin <jdurgin@redhat.com>
925 (cherry picked from commit e7cc6117adf653a4915fb7a75fac68f8fa0239ec)
926
927 commit aef69572588a0dfad58df94cb0d0980d0590d8e4
928 Merge: 69eaad7 534624b
929 Author: Yehuda Sadeh <yehuda@redhat.com>
930 Date: Thu Jan 15 16:40:48 2015 -0800
931
932 Merge branch 'wip-firefly-rgw-backports' into firefly
933
934 commit 534624b7e9decc880e88496355a6fbbe008ede5f
935 Author: Sage Weil <sage@redhat.com>
936 Date: Tue Oct 21 17:59:30 2014 -0700
937
938 init-radosgw.sysv: set ulimit -n before starting daemon
939
940 If we do the ulimit inside the daemon command we will have already
941 dropped privs and will fail.
942
943 Fixes: #9587
944 Backport: giant, firefly
945 Signed-off-by: Sage Weil <sage@redhat.com>
946 (cherry picked from commit 9803cedf54a7baff45ccd0e0f65d2bc220958a46)
947
948 commit fd49cbc535d0d7fa64ebfa458386b47c8ec8616e
949 Author: Yehuda Sadeh <yehuda@redhat.com>
950 Date: Fri Dec 12 17:07:30 2014 -0800
951
952 rgw: use s->bucket_attrs instead of trying to read obj attrs
953
954 Fixes: #10307
955 Backport: firefly, giant
956
957 This is needed, since we can't really read the bucket attrs by trying to
958 read the bucket entry point attrs. We already have the bucket attrs
959 anyway, use these.
960
961 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
962 (cherry picked from commit 5cf193c8686196d5235889e68cb5ea8f1fc8e556)
963
964 commit 79bfffb55a0b3a302368e34417d62f74b82dc224
965 Author: Yehuda Sadeh <yehuda@redhat.com>
966 Date: Wed Nov 5 13:40:55 2014 -0800
967
968 rgw: remove swift user manifest (DLO) hash calculation
969
970 Fixes: #9973
971 Backport: firefly, giant
972
973 Previously we were iterating through the parts, creating hash of the
974 parts etags (as S3 does for multipart uploads). However, swift just
975 calculates the etag for the empty manifest object.
976
977 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
978 (cherry picked from commit ef6d3ad964d34bc526dc4435486bd5c8cdc3b230)
979
980 Conflicts:
981 src/rgw/rgw_op.cc
982
983 commit ac799f0834783590cbb6eb91784c8e0753cb1e03
984 Author: Lei Dong <leidong@yahoo-inc.com>
985 Date: Mon Oct 27 10:29:48 2014 +0800
986
987 fix can not disable max_size quota
988
989 Currently if we enable quota and set max_size = -1, it doesn’t
990 mean max_size is unlimited as expected. Instead, it means object
991 with any size is not allowed to upload because of “QuotaExceeded”.
992 The root cause is the function rgw_rounded_kb which convert max_size
993 to max_size_kb returns 0 for -1 because it takes an unsigned int
994 but we pass an int to it. A simple fix is check max_size before
995 it’s rounded to max_size_kb.
996
997 Test case:
998 1 enable and set quota:
999 radosgw-admin quota enable --uid={user_id} --quota-scope=user
1000 radosgw-admin quota set --quota-scope=user --uid={user_id}\
1001 --max-objects=100 --max-size=-1
1002 2 upload any object with non-zero length
1003 it will return 403 with “QuotaExceeded” and return 200 if you apply the fix.
1004
1005 Fixes: #9907
1006 Backport: giant, firefly
1007 Signed-off-by: Dong Lei leidong@yahoo-inc.com
1008 (cherry picked from commit abd3fd3ef9ee9999b99811937af60b7a5e673e35)
1009
1010 commit 30963fdc98f5650a68b5737729920d43459b5899
1011 Author: Yehuda Sadeh <yehuda@redhat.com>
1012 Date: Tue Sep 23 12:43:55 2014 -0700
1013
1014 rgw: rados->set_attrs() updates bucket index
1015
1016 Fixes: #5595
1017 Backport: dumpling, firefly
1018 We need to update the bucket index when updating object attrs, otherwise
1019 we're missing meta changes that need to be registered. It also
1020 solves issue of bucket index not knowing about object acl changes,
1021 although this one still requires some more work.
1022
1023 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
1024 (cherry picked from commit f833f12a200ecc2c4f74ddb443d6fa61b7ad14db)
1025
1026 commit 31f5e332c616dbb7f7338af3ab37ac65ff66f733
1027 Author: Yehuda Sadeh <yehuda@redhat.com>
1028 Date: Tue Nov 4 22:05:03 2014 -0800
1029
1030 rgw: RGWRados::get_obj() returns wrong len if len == 0
1031
1032 Fixes: #9877
1033 We only updated if len was > 0, should update it if r >= 0. This was the
1034 culprit for issue #9877.
1035 Backport: giant, firefly
1036
1037 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
1038 (cherry picked from commit fe7bf06366adaf787816d1e68f5e3f68e8c91134)
1039
1040 commit 8682d1b15e097ececd927dfbd096dbb46403afca
1041 Author: Yehuda Sadeh <yehuda@redhat.com>
1042 Date: Wed Nov 5 13:28:02 2014 -0800
1043
1044 rgw: send back ETag on S3 object copy
1045
1046 Fixes: #9479
1047 Backport: firefly, giant
1048 We didn't send the etag back correctly. Original code assumed the etag
1049 resided in the attrs, but attrs only contained request attrs.
1050
1051 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
1052 (cherry picked from commit b1bfc3a7e0c9088f01f8ff770ae14f569fbc570d)
1053
1054 Conflicts:
1055 src/rgw/rgw_rados.cc
1056
1057 commit 2d61fc76e97e5e28cf24cc6e341d49a8fe47059c
1058 Author: Yehuda Sadeh <yehuda@redhat.com>
1059 Date: Wed Nov 5 12:35:42 2014 -0800
1060
1061 rgw: S3 object copy content type fix
1062
1063 Fixes: #9478
1064 Backport: firefly, giant
1065 Content type for S3 object copy response should be set to
1066 application/xml.
1067
1068 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
1069 (cherry picked from commit 31963459a0a869c4d32f55baa629f36df33eaa90)
1070
1071 commit 011a8c3bd621337e5a1746f18882be20d3854b14
1072 Author: Yehuda Sadeh <yehuda@redhat.com>
1073 Date: Fri Dec 12 11:20:26 2014 -0800
1074
1075 rgw: http headers need to end with \r\n
1076
1077 Fixes: #9254
1078 Backport: firefly, giant
1079
1080 Reported-by: Benedikt Fraunhofer <fraunhofer@traced.net>
1081 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
1082 (cherry picked from commit 7409ab3df18fb312dd6c9f79084f889c523afdce)
1083
1084 Conflicts:
1085 src/rgw/rgw_civetweb.cc
1086 src/rgw/rgw_fcgi.cc
1087
1088 commit 71d119f2952716aa1f75817e1daaf5fb67ecde94
1089 Author: Yehuda Sadeh <yehuda@redhat.com>
1090 Date: Wed Jan 7 13:56:14 2015 -0800
1091
1092 rgw: index swift keys appropriately
1093
1094 Fixes: #10471
1095 Backport: firefly, giant
1096
1097 We need to index the swift keys by the full uid:subuser when decoding
1098 the json representation, to keep it in line with how we store it when
1099 creating it through other mechanism.
1100
1101 Reported-by: hemant burman <hemant.burman@gmail.com>
1102 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
1103 (cherry picked from commit 478629bd2f3f32afbe6e93eaebb8a8fa01af356f)
1104
1105 commit 7b89a827a063acf78a848a34c1c3298db213950f
1106 Author: Yehuda Sadeh <yehuda@redhat.com>
1107 Date: Thu Nov 20 10:36:05 2014 -0800
1108
1109 rgw-admin: create subuser if needed when creating user
1110
1111 Fixes: #10103
1112 Backport: firefly, giant
1113 This turned up after fixing #9973. Earlier we also didn't create the
1114 subuser in this case, but we didn't really read the subuser info when it
1115 was authenticating. Now we do that as required, so we end up failing the
1116 authentication. This only applies to cases where a subuser was created
1117 using 'user create', and not the 'subuser create' command.
1118
1119 Reviewed-by: Sage Weil <sage@redhat.com>
1120 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
1121 (cherry picked from commit 9ba17a321db06d3d76c9295e411c76842194b25c)
1122
1123 commit 9ee29b5355e9ffeac76707e8d4070bfff5dc99d5
1124 Author: Yehuda Sadeh <yehuda@redhat.com>
1125 Date: Wed Nov 5 14:38:46 2014 -0800
1126
1127 rgw: update swift subuser perm masks when authenticating
1128
1129 Fixes: #9918
1130 Backport: firefly, giant
1131 It seems that we weren't setting the swift perm mask correctly.
1132
1133 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
1134 (cherry picked from commit 5d9f36f757a7272c24d2c9adc31db1ed5e712992)
1135
1136 commit d298fdda2aa177e1a4b43ca94a2292435a50dea4
1137 Author: Yehuda Sadeh <yehuda@redhat.com>
1138 Date: Thu Oct 23 17:39:42 2014 -0700
1139
1140 rgw: send http status reason explicitly in fastcgi
1141
1142 There are issues in certain versions of apache 2.4, where the reason is
1143 not sent back. Instead, just provide the reason explicitly.
1144
1145 Backport: firefly, giant
1146
1147 Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
1148 (cherry picked from commit a9dd4af401328e8f9071dee52470a0685ceb296b)