ceph/doc/changelog/v0.80.5.txt

   1 commit 38b73c67d375a2552d8ed67843c8a65c2c0feba6 (tag: refs/tags/v0.80.5, refs/remotes/gh/last, refs/remotes/gh/firefly)
   2 Author: Jenkins <jenkins@inktank.com>
   3 Date:   Tue Jul 29 05:24:39 2014 -0700
   4
   5     0.80.5
   6
   7 commit b576d5a242c16bc9e38ba283a9784f838614882a
   8 Author: Sage Weil <sage@inktank.com>
   9 Date:   Fri May 9 08:41:33 2014 -0700
  10
  11     osd: cancel agent_timer events on shutdown
  12
  13     We need to cancel all agent timer events on shutdown.  This also needs to
  14     happen early so that any in-progress events will execute before we start
  15     flushing and cleaning up PGs.
  16
  17     Backport: firefly
  18     Signed-off-by: Sage Weil <sage@inktank.com>
  19     (cherry picked from commit c0dc245b662f1f9c640d7dd15fdf4cf26e729782)
  20
  21     Conflicts:
  22         src/osd/OSD.cc
  23
  24 commit 6213130f76b1ee89947bae578a1fccf16b845956
  25 Author: Sage Weil <sage@inktank.com>
  26 Date:   Tue Jul 8 16:11:44 2014 -0700
  27
  28     osd: s/applying repop/canceling repop/
  29
  30     The 'applying' language dates back to when we would wait for acks from
  31     replicas before applying writes locally.  We don't do any of that any more;
  32     now, this loop just cancels the repops with remove_repop() and some other
  33     cleanup.
  34
  35     Signed-off-by: Sage Weil <sage@inktank.com>
  36     (cherry picked from commit ef40737eee4389faa7792661a0f9d15b3d0440f2)
  37
  38 commit 2e7c91587d99c9eba85ab7bd407a5b722b25a81e
  39 Author: Sage Weil <sage@inktank.com>
  40 Date:   Tue Jul 8 16:11:27 2014 -0700
  41
  42     osd: clear PGBackend state on shutdown
  43
  44     This was leaking state on shutdown whenever there were in-flight repops
  45     that were canceled.
  46
  47     Fixes: #7891
  48     Signed-off-by: Sage Weil <sage@inktank.com>
  49     (cherry picked from commit cafceae6c243369d3616af8217884e7e029896b9)
  50
  51 commit f03110f3889ce7340c64f19358840b2425c93f68
  52 Author: Sage Weil <sage@inktank.com>
  53 Date:   Tue Jul 8 16:10:58 2014 -0700
  54
  55     osd: separate cleanup from PGBackend::on_change()
  56
  57     The generic portion of on_change() cleaned up temporary on-disk objects
  58     and requires a Transaction.  The rest is clearing out in-memory state and
  59     does not.  Separate the two.
  60
  61     Signed-off-by: Sage Weil <sage@inktank.com>
  62     (cherry picked from commit e299357e2a24474ab7a02e65ca0279336fd04b76)
  63
  64 commit 42ad241efab751a1d4f50d5a748d91d0fe967844
  65 Merge: 5f4ceb2 5049fc4
  66 Author: Josh Durgin <josh.durgin@inktank.com>
  67 Date:   Wed Jul 23 10:09:19 2014 -0700
  68
  69     Merge remote-tracking branch 'origin/wip-8858-firefly' into firefly
  70
  71     Reviewed-by: Josh Durgin <josh.durgin@inktank.com>
  72
  73 commit 5f4ceb208b55fab7fd1530dcf8fe03198b044148
  74 Author: Joao Eduardo Luis <joao.luis@inktank.com>
  75 Date:   Tue Jul 22 00:25:37 2014 +0100
  76
  77     mon: AuthMonitor: always encode full regardless of keyserver having keys
  78
  79     On clusters without cephx, assuming an admin never added a key to the
  80     cluster, the monitors have empty key servers.  A previous patch had the
  81     AuthMonitor not encoding an empty keyserver as a full version.
  82
  83     As such, whenever the monitor restarts we will have to read the whole
  84     state from disk in the form of incrementals.  This poses a problem upon
  85     trimming, as we do every now and then: whenever we start the monitor, it
  86     will start with an empty keyserver, waiting to be populated from whatever
  87     we have on disk.  This is performed in update_from_paxos(), and the
  88     AuthMonitor's will rely on the keyserver version to decide which
  89     incrementals we care about -- basically, all versions > keyserver version.
  90
  91     Although we started with an empty keyserver (version 0) and are expecting
  92     to read state from disk, in this case it means we will attempt to read
  93     version 1 first.  If the cluster has been running for a while now, and
  94     even if no keys have been added, it's fair to assume that version is
  95     greater than 0 (or even 1), as the AuthMonitor also deals and keeps track
  96     of auth global ids.  As such, we expect to read version 1, then version 2,
  97     and so on.  If we trim at some point however this will not be possible,
  98     as version 1 will not exist -- and we will assert because of that.
  99
 100     This is fixed by ensuring the AuthMonitor keeps track of full versions
 101     of the key server, even if it's of an empty key server -- it will still
 102     keep track of the key server's version, which is incremented each time
 103     we update from paxos even if it is empty.
 104
 105     Fixes: #8851
 106     Backport: dumpling, firefly
 107
 108     Signed-off-by: Joao Eduardo Luis <joao.luis@inktank.com>
 109     (cherry picked from commit b551ae2bcea2dd17b37f5f5ab34251cc78de0e26)
 110
 111 commit 5c1d74848863b45aa9e98d134cbe0cd1c8199557
 112 Merge: 442c3fd c390ec4
 113 Author: Sage Weil <sage@inktank.com>
 114 Date:   Mon Jul 21 13:10:28 2014 -0700
 115
 116     Merge pull request #2122 from ceph/wip-dencoder-firefly
 117
 118     fix linking deps, cpeh-dencoder, etc., firefly backport
 119
 120 commit 5049fc4c9d6513bc58c3c4464f3de18949ac0aa5
 121 Author: Yehuda Sadeh <yehuda@redhat.com>
 122 Date:   Fri Jul 18 14:52:48 2014 -0700
 123
 124     rgw: dump prefix unconditionally
 125
 126     As part of issue #8858, and to be more in line with S3, dump the Prefix
 127     field when listing bucket even if bucket is empty.
 128
 129     Signed-off-by: Yehuda Sadeh <yehuda@redhat.com>
 130     (cherry picked from commit d7209c11251d42227608bc54cc69232ef62ffe80)
 131
 132 commit 7fdd6d94262d8e6c81d168f5b997b3a5f030e6fa
 133 Author: Yehuda Sadeh <yehuda@inktank.com>
 134 Date:   Thu Jul 17 15:48:26 2014 -0700
 135
 136     rgw: list extra objects to set truncation flag correctly
 137
 138     Otherwise we end up returning wrong truncated value, and no data on the
 139     next iteration.
 140
 141     Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
 142     (cherry picked from commit dc417e477d4ad262885c6b5f5987cf06d63b159d)
 143
 144 commit 1f8774e01a616739ab70a78c348ab986b64b8fce
 145 Author: Yehuda Sadeh <yehuda@inktank.com>
 146 Date:   Thu Jul 17 11:45:44 2014 -0700
 147
 148     rgw: account common prefixes for MaxKeys in bucket listing
 149
 150     To be more in line with the S3 api. Beforehand we didn't account the
 151     common prefixes towards the MaxKeys (a single common prefix counts as a
 152     single key). Also need to adjust the marker now if it is pointing at a
 153     common prefix.
 154
 155     Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
 156     (cherry picked from commit 82d2d612e700f94a0bb2d9fb7555abf327be379b)
 157
 158 commit cb0a7ab8eb1d2fc8799b97c5a387b5f4f93b515c
 159 Author: Yehuda Sadeh <yehuda@inktank.com>
 160 Date:   Thu Jul 17 11:24:51 2014 -0700
 161
 162     rgw: add NextMarker param for bucket listing
 163
 164     Partially fixes #8858.
 165
 166     Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
 167     (cherry picked from commit 924686f0b6593deffcd1d4e80ab06b1e7af00dcb)
 168
 169 commit 34aa7086ba5293ee4d008016558dcdb1ada02e83
 170 Author: Yehuda Sadeh <yehuda@inktank.com>
 171 Date:   Wed Jul 16 15:21:09 2014 -0700
 172
 173     rgw: improve delmited listing of bucket
 174
 175     If found a prefix, calculate a string greater than that so that next
 176     request we can skip to that. This is still not the most efficient way to
 177     do it. It'll be better to push it down to the objclass, but that'll
 178     require a much bigger change.
 179
 180     Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
 181     (cherry picked from commit e6cf618c257f26f97f60a4c1df1d23a14496cab0)
 182
 183 commit d027ff1b432a265ad49ddfeafab45093bcd03e6c
 184 Author: Yehuda Sadeh <yehuda@inktank.com>
 185 Date:   Wed Jul 16 16:05:58 2014 -0700
 186
 187     utf8: export encode_utf8() and decode_utf8()
 188
 189     Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
 190     (cherry picked from commit 49fc68cf8c3122c878ea9503c9c74d7046bc9c6f)
 191
 192 commit c390ec4fa919e0a17e646e595b30cfa878c3454f
 193 Author: Sage Weil <sage@redhat.com>
 194 Date:   Fri Jul 18 22:44:51 2014 -0700
 195
 196     ceph-dencoder: don't link librgw.la (and rados, etc.)
 197
 198     Signed-off-by: Sage Weil <sage@redhat.com>
 199     (cherry picked from commit 34671108ce0b7597dba4f619732ecdb8933cda6e)
 200
 201 commit 3d89380c90ab632eb76d3e429788fd81630e7a57
 202 Author: Sage Weil <sage@redhat.com>
 203 Date:   Fri Jul 18 22:27:25 2014 -0700
 204
 205     rgw: move a bunch of stuff into rgw_dencoder
 206
 207     This will help out ceph-dencoder ...
 208
 209     Signed-off-by: Sage Weil <sage@redhat.com>
 210     (cherry picked from commit b1a641f307942cbf43036f75ef67fb30441dfe95)
 211
 212 commit dc6c9bf2b92d28bbd2bc822c267b0555f74c87fc
 213 Author: Sage Weil <sage@redhat.com>
 214 Date:   Fri Jul 18 21:58:29 2014 -0700
 215
 216     libosd_types, libos_types, libmon_types
 217
 218     Signed-off-by: Sage Weil <sage@redhat.com>
 219     (cherry picked from commit 1c170776cb8c0266f0f54f049ed07bbdb9c9ab5e)
 220
 221     Conflicts:
 222
 223         src/os/Makefile.am
 224         src/os/ObjectStore.cc
 225         src/osd/Makefile.am
 226
 227 commit 7be91e8c5d935f7094033842bde1da2486905c70
 228 Author: Sage Weil <sage@redhat.com>
 229 Date:   Fri Jul 18 20:55:39 2014 -0700
 230
 231     Revert "ceph.spec: move ceph-dencoder to ceph from ceph-common"
 232
 233     This reverts commit 95f5a448b52db545a2b9bbad47fdb287254f93ea.
 234     (cherry picked from commit 58cc894b3252a848ebc2169bcc4980a0ae6cc375)
 235
 236 commit 04ad08a18a773e104a826b2a77ab0b2ffaac1e5f
 237 Author: Sage Weil <sage@redhat.com>
 238 Date:   Fri Jul 18 20:55:35 2014 -0700
 239
 240     Revert "debian: move ceph-dencoder to ceph from ceph-common"
 241
 242     This reverts commit b37e3bde3bd31287b11c069062280258666df7c5.
 243     (cherry picked from commit f181f78b7473260a717bc8ab4fc4d73a80e3b5ba)
 244
 245 commit 442c3fd5a3b6a6c6a85d9bd7755923f14aa598e7
 246 Author: Dan Mick <dan.mick@inktank.com>
 247 Date:   Thu Jul 17 21:44:06 2014 -0700
 248
 249     configure: do not link leveldb with everything
 250
 251     Detect leveldb, but do not let autoconf blindly link it with everything on the
 252     planet.
 253
 254     Signed-off-by: Dan Mick <dan.mick@inktank.com>
 255     Sighed-off-by: Sage Weil <sage@redhat.com>
 256     (cherry picked from commit 06a8f7b99c5533f397b34f448138220384df60ac)
 257
 258 commit c1073e430eafeb1f213f80636a3a24966b292698
 259 Author: Sage Weil <sage@redhat.com>
 260 Date:   Mon Jul 14 17:02:05 2014 -0700
 261
 262     ceph.spec: move ceph-dencoder to ceph from ceph-common
 263
 264     Signed-off-by: Sage Weil <sage@redhat.com>
 265     (cherry picked from commit 95f5a448b52db545a2b9bbad47fdb287254f93ea)
 266
 267 commit f4e9e0ef2eab3174584f1695122b114549feff57
 268 Author: Sage Weil <sage@redhat.com>
 269 Date:   Mon Jul 14 14:16:43 2014 -0700
 270
 271     debian: move ceph-dencoder to ceph from ceph-common
 272
 273     It links against the world currently (notably leveldb).  Not nice for the
 274     client-side lib.
 275
 276     Signed-off-by: Sage Weil <sage@redhat.com>
 277     (cherry picked from commit b37e3bde3bd31287b11c069062280258666df7c5)
 278
 279 commit 681c399e83df4c6ebd7e0836203eb2466455f548
 280 Author: Sage Weil <sage@redhat.com>
 281 Date:   Thu Jul 17 16:40:06 2014 -0700
 282
 283     logrotate.conf: fix osd log rotation under upstart
 284
 285     In commit 7411c3c6a42bef5987bdd76b1812b01686303502 we generalized this
 286     enumeration code by copying what was in the upstart scripts.  However,
 287     while the mon and mds directories get a 'done' file, the OSDs get a 'ready'
 288     file.  Bah!  Trigger off of either one.
 289
 290     Backport: firefly
 291     Signed-off-by: Sage Weil <sage@redhat.com>
 292     (cherry picked from commit 14a9ca602fa9573b0b6b94709b384bca02d12fac)
 293
 294 commit 794ec7fb43da4db930d28571a84c1c1cec1eb066
 295 Author: Yehuda Sadeh <yehuda@inktank.com>
 296 Date:   Wed Jul 16 12:23:31 2014 -0700
 297
 298     rgw: don't try to wait for pending if list is empty
 299
 300     Fixes: #8846
 301     Backport: firefly, dumpling
 302
 303     This was broken at ea68b9372319fd0bab40856db26528d36359102e. We ended
 304     up calling wait_pending_front() when pending list was empty.
 305     This commit also moves the need_to_wait check to a different place,
 306     where we actually throttle (and not just drain completed IOs).
 307
 308     Reported-by: Sylvain Munaut <s.munaut@whatever-company.com>
 309     Signed-off-by: Yehuda Sadeh <yehuda@inktank.com>
 310     (cherry picked from commit f9f2417d7db01ecf2425039539997901615816a9)
 311
 312 commit e75dd2e4b7adb65c2de84e633efcd6c19a6e457b
 313 Author: Alfredo Deza <alfredo.deza@inktank.com>
 314 Date:   Thu Jul 17 10:06:37 2014 -0400
 315
 316     remove suse service restarts
 317
 318     Signed-off-by: Alfredo Deza <alfredo.deza@inktank.com>
 319     (cherry picked from commit 0f11aaeefd7e8b49f88607937aade6113ebda52c)
 320
 321 commit 361c1f8554ce1fedfd0020cd306c41b0ba25f53e
 322 Author: Alfredo Deza <alfredo.deza@inktank.com>
 323 Date:   Wed Jul 16 16:12:12 2014 -0400
 324
 325     remove ceph restarts on upgrades for RPMs
 326
 327     Signed-off-by: Alfredo Deza <alfredo.deza@inktank.com>
 328     (cherry picked from commit e3a5756f58ef5d07badf18ab08a26f47f7d232cb)
 329
 330 commit 88f694dcd3d3342f764f3790572988b7808028ae
 331 Author: Alfredo Deza <alfredo.deza@inktank.com>
 332 Date:   Wed Jul 16 11:07:13 2014 -0400
 333
 334     set the default log level to WARNING
 335
 336     Signed-off-by: Alfredo Deza <alfredo.deza@inktank.com>
 337     (cherry picked from commit 420f0a471a31d1e56359ea642ed831e8a38b1e90)
 338
 339 commit 67b5193f73a2c9ec9e503ad3431473998217375d
 340 Author: Sage Weil <sage@redhat.com>
 341 Date:   Tue Jul 15 18:11:41 2014 -0700
 342
 343     init-ceph: wrap daemon startup with systemd-run when running under systemd
 344
 345     We want to make sure the daemon runs in its own systemd environment.  Check
 346     for systemd as pid 1 and, when present, use systemd-run -r <cmd> to do
 347     this.
 348
 349     Probably fixes #7627
 350
 351     Signed-off-by: Sage Weil <sage@redhat.com>
 352     Reviewed-by: Dan Mick <dan.mick@inktank.com>
 353     Tested-by: Dan Mick <dan.mick@inktank.com>
 354     (cherry picked from commit 3e0d9800767018625f0e7d797c812aa44c426dab)