commit ba340a97c3dafc9155023da8d515eecc675c619a (tag: refs/tags/v0.67.9, refs/remotes/gh/dumpling) Author: Jenkins Date: Wed May 21 16:57:02 2014 +0000 0.67.9 commit b638d19d126646d2a8f6da11067c5f392a62525e Author: Guang Yang Date: Fri May 9 09:21:23 2014 +0000 msg: Fix inconsistent message sequence negotiation during connection reset Backport: firefly, emperor, dumpling Signed-off-by: Guang Yang (yguang@yahoo-inc.com) Reviewed-by: Greg Farnum (cherry picked from commit bdee119076dd0eb65334840d141ccdf06091e3c9) commit 55b1112819ed2e6fd509c09c2c1a2b434d22aac6 Author: Sage Weil Date: Tue May 20 10:46:34 2014 -0700 OSD::handle_pg_query: on dne pg, send lb=hobject_t() if deleting We will set lb=hobject_t() if we resurrect the pg. In that case, we need to have sent that to the primary before hand. If we finish the removal before the pg is recreated, we'll just end up backfilling it, which is ok since the pg doesn't exist anyway. Fixes: #7740 Signed-off-by: Samuel Just (cherry picked from commit 04de781765dd5ac0e28dd1a43cfe85020c0854f8) Conflicts: src/osd/OSD.cc commit 3b8ab41e1ec86f2ab5c6b4bee3fb4030077e2c21 Author: Sage Weil Date: Fri May 2 14:48:35 2014 -0700 mon/MonClient: remove stray _finish_hunting() calls Callig _finish_hunting() clears out the bool hunting flag, which means we don't retry by connection to another mon periodically. Instead, we send keepalives every 10s. But, since we aren't yet in state HAVE_SESSION, we don't check that the keepalives are getting responses. This means that an ill-timed connection reset (say, after we get a MonMap, but before we finish authenticating) can drop the monc into a black hole that does not retry. Instead, we should *only* call _finish_hunting() when we complete the authentication handshake. Fixes: #8278 Backport: firefly, dumpling Signed-off-by: Sage Weil Reviewed-by: Joao Eduardo Luis (cherry picked from commit 77a6f0aefebebf057f02bfb95c088a30ed93c53f) commit a915350491092abcd2584d0bbb5f5d8610ed89fc Merge: bd5d6f1 d4e67ff Author: Sage Weil Date: Tue May 20 10:19:00 2014 -0700 Merge pull request #1826 from ceph/wip-8162-dumpling Wip 8162 dumpling Reviewed-by: Sage Weil commit bd5d6f116416d1b410d57ce00cb3e2abf6de102b Author: Greg Farnum Date: Thu May 15 16:50:43 2014 -0700 OSD: fix an osdmap_subscribe interface misuse When calling osdmap_subscribe, you have to pass an epoch newer than the current map's. _maybe_boot() was not doing this correctly -- we would fail a check for being *in* the monitor's existing map range, and then pass along the map prior to the monitor's range. But if we were exactly one behind, that value would be our current epoch, and the request would get dropped. So instead, make sure we are not *in contact* with the monitor's existing map range. Signed-off-by: Greg Farnum Reviewed-by: Samuel Just (cherry picked from commit 290ac818696414758978b78517b137c226110bb4) commit c5ae8ffc281ac538dd0966a552a6cf729728e925 Merge: fced056 8d991fa Author: Sage Weil Date: Mon May 19 13:45:50 2014 -0700 Merge pull request #1827 from ceph/wip-6565-dumpling Wip 6565 dumpling Reviewed-by: Sage Weil commit 8d991fa2c3fdbfd9178d93e4ca33d47504a37195 Author: Samuel Just Date: Wed Oct 16 10:07:37 2013 -0700 OSD: check for splitting when processing recover/backfill reservations Fixes: 6565 Signed-off-by: Samuel Just Reviewed-by: Sage Weil (cherry picked from commit 15ec5332ba4154930a0447e2bcf1acec02691e97) commit d4e67ff3037a3cc7ae2ecc9e1d8d086c45ae515a Author: Samuel Just Date: Thu May 8 13:25:32 2014 -0700 ReplicatedPG::recover_backfill: do not update last_backfill prematurely Previously, we would update last_backfill on the backfill peer to backfills_in_flight.empty() ? backfill_pos : backfills_in_flight.begin()->first which is actually the next backfill to complete. We want to update last_backfill to the largest completed backfill instead. We use the pending_backfill_updates mapping to identify the most recently completed backfill. Due to the previous patch, deletes will also be included in that mapping. Related sha1s from master: 4139e75d63b0503dbb7fea8036044eda5e8b7cf1 7a06a71e0f2023f66d003dfb0168f4fe51eaa058 We don't really want to backport those due to the changes in: 9ec35d5ccf6a86c380865c7fc96017a1f502560a This patch does essentially the same thing, but using backfill_pos. Fixse: #8162 Signed-off-by: Samuel Just commit d620b13c51fc69b509824177af32de55c89c2259 Author: Samuel Just Date: Mon Oct 28 16:03:25 2013 -0700 ReplicatedPG: add empty stat when we remove an object in recover_backfill Subsequent updates to that object need to have their stats added to the backfill info stats atomically with the last_backfill update. Signed-off-by: Samuel Just (cherry picked from commit ecddd12b01be120fba87f5ac60539f98f2c69a28) commit fced0562f6b80035cd522cc81bdc844ccc92d9bc Author: Yehuda Sadeh Date: Wed Nov 27 13:34:00 2013 -0800 rgw: don't error out on empty owner when setting acls Fixes: #6892 Backport: dumpling, emperor s3cmd specifies empty owner field when trying to set acls on object / bucket. We errored out as it didn't match the current owner name, but with this change we ignore it. Signed-off-by: Yehuda Sadeh (cherry picked from commit 14cf4caff58cc2c535101d48c53afd54d8632104) commit 0e685c68aa6a500aa7fa433cd9b8246f70c5383e Author: Yehuda Sadeh Date: Mon Apr 21 15:34:04 2014 -0700 rgw: send user manifest header field Fixes: #8170 Backport: firefly If user manifest header exists (swift) send it as part of the object header data. Signed-off-by: Yehuda Sadeh (cherry picked from commit 5cc5686039a882ad345681133c9c5a4a2c2fd86b) commit 20383e359cdfbf146838742fa050ec4b6073271d Author: Yan, Zheng Date: Fri Apr 11 15:03:37 2014 +0800 client: add asok command to kick sessions that were remote reset Fixes: #8021 Signed-off-by: Yan, Zheng (cherry picked from commit 09a1bc5a4601d356b9cc69be8541e6515d763861) commit 4e5e41deeaf91c885773d90e6f94da60f6d4efd3 Author: Sage Weil Date: Fri Apr 18 13:50:11 2014 -0700 osd: throttle snap trimmming with simple delay This is not particularly smart, but it is *a* knob that lets you make the snap trimmer slow down. It's a flow and a simple delay, so it is adjustable at runtime. Default is 0 (no change in behavior). Partial solution for #6278. Signed-off-by: Sage Weil (cherry picked from commit 4413670d784efc2392359f0f22bca7c9056188f4) commit 14650b282ecd344025714a4e743b255ae01b3ce0 Author: Sage Weil Date: Tue Apr 1 16:01:28 2014 -0700 PG: only complete replicas should count toward min_size Backport: emperor,dumpling,cuttlefish Fixes: #7805 Signed-off-by: Samuel Just Signed-off-by: Sage Weil (cherry picked from commit 0d5d3d1a30685e7c47173b974caa12076c43a9c4) commit 3f1d7f5e0a67ad646de465335fb7ee00eb07e220 Author: Sage Weil Date: Fri May 2 15:10:43 2014 -0700 mon/PGMonitor: set tid on no-op PGStatsAck The OSD needs to know the tid. Both generally, and specifically because the flush_pg_stats may be blocking on it. Fixes: #8280 Signed-off-by: Sage Weil Reviewed-by: Joao Eduardo Luis (cherry picked from commit 5a6ae2a978dcaf96ef89de3aaa74fe951a64def6)