RepoMirrors/ceph
1
0
Fork 0
mirror of https://github.com/ceph/ceph synced 2024-12-18 09:25:49 +00:00
Code Issues Packages Projects Releases Wiki Activity
68,463 Commits 211 Branches 394 Tags 2.1 GiB
a6fa7b6568
Commit Graph

68463 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Sage Weil
a6fa7b6568 osd/OSDMap: generalize map_to_pg 
So we can do this without constructing an object_locator_t.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-13 23:03:50 -05:00
Sage Weil
0efdd0a338 osd: make use of MOSDOp::get_hobj() 
Prefer this to get_object_locator() whereever possible.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-13 23:03:50 -05:00
Sage Weil
e0c037e199 message/MOSDOp: build native hobject_t 
Drop unneeded snapid_t snapid and object_locator_t, which just duplicate
hobject_t fields.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-13 23:03:49 -05:00
Sage Weil
e9cfeedb5d osd/PG: fix tracking of last_epoch_split 
Note that it is only (currently) important that this value be accurate
on the current OSD since we only use this value (currently) to discard
ops sent before the split.  If we are getting the history from a different
OSD in the cluster that doesn't have an up to date value it doesn't matter
because that implies a primary change and also a client resend.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-13 23:03:49 -05:00
Sage Weil
36d3a29ef5 osd/PG: discard ops from before the last split 
New clients will resend.

Old clients will see a last_force_op_resend (now named
last_force_op_resend_preluminous in latest code) and resend.

We know this because we require that the monitors upgrade to luminous
before the OSDs, and the new mon code sets this field on split.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-13 23:03:49 -05:00
Sage Weil
6e65e2665d osdc/Objecter: resend ops on pg split if osd has CEPH_FEATURE_RESEND_ON_SPLIT 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-13 23:03:49 -05:00
Sage Weil
189ca7ec64 mon/OSDMonitor: make pre-luminous clients resend ops on split 
Pre-luminous clients do not understand that a split PG forms a new
interval.  Make them resend ops to work around this.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-13 11:44:53 -05:00
Sage Weil
8caa8f162e osd/PG: discard ops based on either new or old lfor and features 
If the client has the new feature bit, use the new field; if they have the
older feature bit, use the old field.

Note that there is no change to the Objecter: last_force_op_resend is
still the "current" field that it should pay attention to.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-13 11:44:53 -05:00
Sage Weil
0423b541b3 osd/osd_types: last_force_op_resend -> last_force_op_resend_preluminous 
Rename the current last_force_op_resend for legacy clients, and add a new
one that only applies to new clients that have the new
CEPH_FEATURE_OSD_NEW_INTERVAL_ON_SPLIT feature.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-13 11:44:53 -05:00
Sage Weil
6de5481415 osd/osd_types: add set_last_force_op_resend() accessor and use it 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-13 11:44:53 -05:00
Sage Weil
316dc98788 Merge pull request #13321 from liewegas/wip-kill-sortbitwise-harder 
osd: kill sortbitwise

Reviewed-by: Brad Hubbard <bhubbard@redhat.com>
 2017-02-13 10:43:56 -06:00
Matt Benjamin
24f43e9f51 Merge pull request #13303 from Jing-Scott/fix-swift-cannot-disable-object-versioning 
rgw: fix swift cannot disable object versioning

@Jing-Scott  updated, addressing @rzarzynski's change request
 2017-02-13 10:58:04 -05:00
Kefu Chai
1383652b93 Merge pull request #13393 from ddiss/fix_rgw_fcgi_include 
rgw: set FCGI_INCLUDE_DIR for cephd_rgw_base

Reviewed-by: Casey Bodley <cbodley@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>
 2017-02-13 23:28:03 +08:00
Sage Weil
3ea49b2422 Merge pull request #13382 from xiexingguo/wip-counter-name 
os/bluestore: fix wal-queue bytes-counter to keep pace with others

Reviewed-by: Sage Weil <sage@redhat.com>
 2017-02-13 08:25:51 -06:00
David Disseldorp
3e55db45f9 rgw: set FCGI_INCLUDE_DIR for cephd_rgw_base 
cephd_rgw_base build currently fails with fastcgi enabled:
--
In file included from /home/david/ceph/src/rgw/rgw_request.h:13:0,
from /home/david/ceph/src/rgw/rgw_main.cc:53:
/home/david/ceph/src/rgw/rgw_fcgi.h:8:21: fatal error: fcgiapp.h:
					No such file or directory
^
--

This is despite the fact that fastcgi was detected and located at
configure time:
build/CMakeCache.txt:FCGI_INCLUDE_DIR:PATH=/usr/include/fastcgi

Fix this by ensuring that the cephd_rgw_base build target correctly uses
FCGI_INCLUDE_DIR.

Fixes: http://tracker.ceph.com/issues/18918

Signed-off-by: David Disseldorp <ddiss@suse.de>
 2017-02-13 15:03:42 +01:00
Haomai Wang
d914f0f8a6 Merge pull request #13392 from Adirl/zero_inline 
msg/async/rdma: zero wqe inline

Reviewed-by: Haomai Wang <haomai@xsky.com>
 2017-02-13 21:57:19 +08:00
Adir Lev
442e22e24f msg/async/rdma: zero wqe inline 
we dont use wqe inline

issue: none

Change-Id: Ib187eec3bc79f8caf040fbbfbf0a915186cd58ba
Signed-off-by: Adir Lev <adirl@mellanox.com>
 2017-02-13 13:36:46 +00:00
Jing Wenjun
17c5a0edd2 rgw: fix swift cannot disable object versioning 
we should be able to disable object verioning by removing its X-Versions-Location
metadata header by sending an empty key value. this description can be found at
No.8 in http://docs.openstack.org/user-guide/cli-swift-set-object-versions.html.

Fixes: http://tracker.ceph.com/issues/18852
Signed-off-by: Jing Wenjun <jingwenjun@cmss.chinamobile.com>
 2017-02-13 21:13:05 +08:00
Kefu Chai
877c938937 Merge pull request #13380 from tchaikov/wip-sepia-bt 
script/sepia_bt.sh: no need to pass version and sha1 anymore

Reviewed-by: Brad Hubbard <bhubbard@redhat.com>
 2017-02-13 20:48:17 +08:00
Nathan Cutler
6394d2fd6c Merge pull request #13368 from smithfarm/wip-doc-dev 
doc: instructions for provisioning OpenStack VMs ad hoc

Reviewed-by: Loic Dachary <ldachary@redhat.com>
 2017-02-13 12:39:59 +01:00
Kefu Chai
148d48812c Merge pull request #13367 from tchaikov/wip-qa-jewel-x-singleton 
test: Thrasher: update pgp_num of all expanded pools if not yet

Reviewed-by: Sage Weil <sage@redhat.com>
 2017-02-13 14:54:27 +08:00
Kefu Chai
dbb6195662 script/sepia_bt.sh: no need to pass version and sha1 anymore 
* try to figure out the working directory from the core dump's path
* add -C directory option, so user is able to specify the directory
  for the downloaded packages, and extracted files.

Signed-off-by: Kefu Chai <kchai@redhat.com>
 2017-02-13 14:33:18 +08:00
xie xingguo
f50990d544 os/bluestore: fix wal-queue bytes-counter to keep pace with others 
Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
 2017-02-13 14:09:45 +08:00
Kefu Chai
de59b5102c test: Thrasher: restore changed options after done with thrash 
Signed-off-by: Kefu Chai <kchai@redhat.com>
 2017-02-13 09:25:51 +08:00
Kefu Chai
761a1dc391 tests: Thrasher: extract _set_config() method 
Signed-off-by: Kefu Chai <kchai@redhat.com>
 2017-02-13 09:25:50 +08:00
Kefu Chai
995e144e3e tests: CephManager: add get_config() method 
Signed-off-by: Kefu Chai <kchai@redhat.com>
 2017-02-13 09:25:50 +08:00
Kefu Chai
136483a8f9 test: Thrasher: update pgp_num of all expanded pools if not yet 
otherwise wait_until_healthy will fail after timeout as seeing warning
like:

HEALTH_WARN pool cephfs_data pg_num 182 > pgp_num 172

Signed-off-by: Kefu Chai <kchai@redhat.com>
 2017-02-13 09:25:50 +08:00
Yuri Weinstein
05f1a1987e Merge pull request #13378 from liewegas/wip-luminous 
qa/rados/upgrade/jewel-x-singleton: run luminous.yaml at the end

Reviewed-by: Yuri Weinstein <yweins@redhat.com>
 2017-02-12 08:13:26 -08:00
Sage Weil
5f6384db76 Merge pull request #13365 from xiexingguo/wip-backoff-refleak 
osdc/Objecter: fix possible OSDSession leak on wrong connection

Reviewed-by: Sage Weil <sage@redhat.com>
 2017-02-12 08:50:38 -06:00
Sage Weil
cfccde7aaf Merge pull request #13363 from jdurgin/wip-req-luminous-error 
osd: clarify REQUIRE_LUMINOUS error message

Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>
Reviewed-by: Kefu Chai <kchai@redhat.com>
 2017-02-12 08:49:57 -06:00
Sage Weil
6e998b2f8c qa/rados/upgrade/jewel-x-singleton: run luminous.yaml at the end 
renumber too

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-12 09:47:41 -05:00
Nathan Cutler
63100b4f9b doc: instructions for provisioning OpenStack VMs ad hoc 
Signed-off-by: Nathan Cutler <ncutler@suse.com>
 2017-02-11 18:57:33 +01:00
Sage Weil
784c7edb4e osd/PrimaryLogPG: unconditionally set PGNLS next to hobject 
Drop nibblewise workaround.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-11 10:45:17 -05:00
Sage Weil
75d7a5c85f osd/PG: drop get_sort_bitwise() accessors, flag 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-11 10:45:17 -05:00
Sage Weil
b02439ff1c osd/PG: always bitwise 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-11 10:45:16 -05:00
Sage Weil
d55cd22150 osd/PG: drop bitwise arg to MissingLoc::rebuild() 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-11 10:45:16 -05:00
Sage Weil
c5f880443d osd/osd_types: make set_last_backfill() assume bitwise 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-11 10:45:16 -05:00
Sage Weil
32acb10bd0 osd: BackfillInterval is always bitwise 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-11 10:45:16 -05:00
Sage Weil
10aab91f5e osd: fix add_source_info 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-11 10:45:16 -05:00
Sage Weil
ad4365b735 osd: fix write_blocked_by_scrub 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-11 10:45:16 -05:00
Sage Weil
b052e497a5 osd/osd_types: drop bitwise from ScrubMap 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-11 10:45:16 -05:00
Sage Weil
1699bf7b71 common/hobject: eliminate wonky compartors 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-11 10:45:16 -05:00
Sage Weil
43244da571 common/hobject: remove cmp_* comparators; add normal operators 
Fix up callers.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-11 10:45:16 -05:00
Sage Weil
e8c55fe157 os: remove nibblewise sort order support 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-11 10:45:16 -05:00
Sage Weil
ea8b17035e osd: shutdown if sortbitwise flag is ever cleared 
Note we do this only in activate_map so that we can (while booting)
skip past maps that don't have it set.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-02-11 10:45:16 -05:00
Ilya Dryomov
eb8d3372e8 Merge pull request #13369 from idryomov/wip-doc-kernel-vers 
doc: kernel client os-recommendations update

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
 2017-02-11 16:01:02 +01:00
Jason Dillaman
e95fe0dfad Merge pull request #13229 from trociny/wip-18335 
rbd-nbd: check /sys/block/nbdX/size to ensure kernel mapped correctly

Reviewed-by: Jason Dillaman <dillaman@redhat.com>
 2017-02-11 09:29:23 -05:00
Kefu Chai
54354e597e Merge pull request #13292 from tchaikov/wip-script-sepia-bt-support-xenial 
script/sepia_bt.sh: support xenial

Reviewed-by: Brad Hubbard <bhubbard@redhat.com>
 2017-02-11 12:22:34 +08:00
xie xingguo
ad304a7d8e osdc/Objecter: fix possible OSDSession leak on wrong connection 
This is introduced by the newly added backoff logic.
Not sure if it will really happen, but just in case.

Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
 2017-02-11 11:01:15 +08:00
Sage Weil
d4db8e04dd Merge pull request #12342 from liewegas/wip-backoff 
osd,osdc: pg and osd-based backoff

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
 2017-02-10 19:47:49 -06:00