Commit Graph

75979 Commits

Author SHA1 Message Date
John Spray
48aa368db9 common: reinstate freebsd ifdefs in options.cc
The automated conversion missed these.

Signed-off-by: John Spray <john.spray@redhat.com>
2017-07-24 09:04:00 -04:00
John Spray
4d06719f10 common: use code_environment_t for daemon-ness in config
...and reinstate the set_daemon_default versions of config options
in place of the special casing in common_init.

Signed-off-by: John Spray <john.spray@redhat.com>
2017-07-24 09:03:59 -04:00
John Spray
89acbd5310 common/options: mark all obvious LEVEL_DEV options
This is everthing matching kill.*at, debug, inject.

Signed-off-by: John Spray <john.spray@redhat.com>
2017-07-24 09:03:59 -04:00
John Spray
69eaa7ff9f common: separate out options by subsystem
This avoids compiler warning about one oversized
literal, allows us to add_service en-masse to
the whole collection, and avoids potential merge
conflicts when config changes in separate subsystems
are made.

Signed-off-by: John Spray <john.spray@redhat.com>
2017-07-24 09:03:59 -04:00
John Spray
34f412573d Merge pull request #16360 from dillaman/wip-rbd-mirror-service-daemon
mgr/dashboard: rbd mirroring status page

Reviewed-by: John Spray <john.spray@redhat.com>
2017-07-24 14:00:51 +01:00
John Spray
478ba0568e Merge pull request #16529 from Liuchang0812/wip-20750
mgr/status: row has incorrect number of values

Reviewed-by: John Spray <john.spray@redhat.com>
2017-07-24 12:40:52 +01:00
Kefu Chai
fa36dc3e55 Merge pull request #16503 from renhwztetecs/renhw-wip-mon-rebuild-doc
doc/mon: fix ceph-authtool command in rebuild mon's sample

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-07-24 19:39:44 +08:00
chenhg
5dd3dff0f6 OSDMonitor: check mon_max_pool_pg_num when set pool pg_num
Check the mon_max_pool_pg_num when creating pools, but did not

check in the the modification. So we can create a pool and then

modify pg_num exceeds the mon_max_pool_pg_num limit.

By following these steps can verify this problem

step 1: set  mon_max_pool_pg_num = 65536

step 2: ceph osd pool create testpool 65536 65536

step 3: ceph osd pool set testpool pg_num 65537

Signed-off-by: chenhg <c744402859@gmail.com>
2017-07-24 18:36:30 +08:00
Kefu Chai
8cf3b6b92f Merge pull request #16523 from tchaikov/wip-ceph-disk-asok-dir
tests: add setup/teardown for asok dir

Reviewed-by: Loic Dachary <ldachary@redhat.com>
Reviewed-by: Willem Jan Withagen <wjw@digiware.nl>
2017-07-24 18:31:37 +08:00
Kefu Chai
e7e58277a6 doc/rados/operations/add-or-rm-osds: fix dead hyper link
Signed-off-by: Kefu Chai <kchai@redhat.com>
2017-07-24 18:16:25 +08:00
Kefu Chai
780c9acd43 doc: document "ceph osd set-require-min-compat-client"
Signed-off-by: Kefu Chai <kchai@redhat.com>
2017-07-24 18:16:25 +08:00
liuchang0812
c7a88e42d1 mgr/status: row has incorrect number of values
Fixes: http://tracker.ceph.com/issues/20750

Signed-off-by: liuchang0812 <liuchang0812@gmail.com>
2017-07-24 17:34:45 +08:00
Xie Xingguo
42d14a62c8 Merge pull request #16507 from scienceluo/patch-1
Update .organizationmap

Reviewed-by: Jos Collin <jcollin@redhat.com>
Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>
2017-07-24 17:28:24 +08:00
huanwen ren
f4dd2801ca doc/mon: fix ceph-authtool create keyring
if use "ceph-authtool keyring -n mon. --cap mon allow 'allow *'"
as err:
[root]# ceph-authtool keyring -n mon. --cap mon allow 'allow *'
ceph-authtool: unexpected 'allow *'

Excessive "allow" is the cause of the error

Signed-off-by huanwen ren <ren.huanwen@zte.com.cn>
2017-07-24 17:00:10 +08:00
Kefu Chai
4a6a3413b2 doc: document "ceph features"
Signed-off-by: Kefu Chai <kchai@redhat.com>
2017-07-24 16:54:10 +08:00
Kefu Chai
8c1a2b1d94 vstart.sh: do not try to remove default asok_dir
Signed-off-by: Kefu Chai <kchai@redhat.com>
2017-07-24 16:40:10 +08:00
Kefu Chai
b0bd816dd1 tests: add setup/teardown for asok dir
ceph-disk.sh has its own setup/teardown, so update them also. so asok
files will have their parent directory when being created.

Signed-off-by: Kefu Chai <kchai@redhat.com>
2017-07-24 16:39:17 +08:00
Orit Wasserman
7536d2ecae Merge pull request #16502 from weiqiaomiao/wip_weiqm_unlock_bucket
rgw: should unlock when reshard_log->update() reture non-zero in RGWB…
Reviewed-by: Orit Wasserman <owasserm@redhat.com>
2017-07-24 11:25:21 +03:00
weiqiaomiao
86c8b720a1 rgw: Drop Unused function in rgw_usage.cc
fix compile warnning in any PR (Ex: https://jenkins.ceph.com/job/ceph-pull-requests/29193/console):
/home/jenkins-build/build/workspace/ceph-pull-requests/src/rgw/rgw_usage.cc:33:13: warning: ‘void dump_usage_bucket_info(ceph::Formatter*, std::__cxx11::string&, cls_user_bucket_entry&)’ defined but not used [-Wunused-function]
 static void dump_usage_bucket_info(Formatter *formatter, string& name, cls_user_bucket_entry& entry)

Signed-off-by: Wei Qiaomiao <wei.qiaomiao@zte.com.cn>
2017-07-24 12:32:00 +08:00
weiqiaomiao
ff5bd5bd52 rgw: should unlock when reshard_log->update() reture non-zero in RGWBucketReshard::execute()
Signed-off-by: Wei Qiaomiao <wei.qiaomiao@zte.com.cn>
2017-07-24 12:30:01 +08:00
Sage Weil
23b6df4ef0 osd/OSDMap: fix Incremental::dump
Signed-off-by: Sage Weil <sage@redhat.com>
2017-07-23 22:03:49 -04:00
Sage Weil
a1611c89df mon/OSDMonitor: remove zeroed new_state updates
These are interpreted as CEPH_OSD_UP in apply_incremental for legacy
reasons--we don't want them!

Fixes: http://tracker.ceph.com/issues/20751
Signed-off-by: Sage Weil <sage@redhat.com>
2017-07-23 22:03:47 -04:00
Pan Liu
67991e1f6f common: the latency dumped by "ceph osd perf" is not real
Signed-off-by: Pan Liu <wanjun.lp@alibaba-inc.com>
2017-07-23 15:55:47 +08:00
Josh Durgin
3f591d4f43 osd: PrimaryLogPG, PGBackend: complete callback even if interval changes
This avoids leaking the callback passed to remove_missing_object().
Use EAGAIN to signal that the work was not completed, so we don't have
to take the pg lock and check if the interval changed again in
PGBackend.

Fixes: http://tracker.ceph.com/issues/20747
Signed-off-by: Josh Durgin <jdurgin@redhat.com>
2017-07-23 02:27:53 -04:00
Sage Weil
27c1d62ae9 Merge pull request #16508 from liewegas/wip-mgr-lock-cycle
mgr: fix lock cycle
2017-07-22 21:14:36 -05:00
Sage Weil
ecd1193ab9 qa/suites/rados/basic/tasks/rgw_snaps: wait for pools to be be created
Signed-off-by: Sage Weil <sage@redhat.com>
2017-07-22 18:54:46 -04:00
Sage Weil
9b4002b6b8 qa/suites/rados/basic/tasks/rgw_snaps: fix pool list
Signed-off-by: Sage Weil <sage@redhat.com>
2017-07-22 18:54:45 -04:00
Sage Weil
17c1a8f74a Merge pull request #16488 from liewegas/wip-skewed
common/options: remove mon_warn_osd_usage_min_max_delta from options.cc too

Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>
2017-07-22 13:07:43 -05:00
Sage Weil
1fbf4ca7fd Merge pull request #16490 from jdurgin/wip-20739
osd/PrimaryLogPG: skip deleted missing objects in pg[n]ls

Reviewed-by: Sage Weil <sage@redhat.com>
2017-07-22 13:07:28 -05:00
Sage Weil
a7b40ce13c Merge pull request #16501 from liewegas/wip-fix-mgr-commands
mon: do not dereference empty mgr_commands
2017-07-22 13:07:05 -05:00
Sage Weil
08bdc2c867 Merge pull request #16500 from liewegas/wip-compact-sudo
qa/workunits/cephtool/test.sh: add sudo for daemon compact

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-07-22 13:01:20 -05:00
Sage Weil
bb2c84eda6 mgr/PyModules: fix lock name
Signed-off-by: Sage Weil <sage@redhat.com>
2017-07-22 13:48:17 -04:00
Sage Weil
d7ba56f421 mgr/DaemonServer: fix lock cycle
This breaks a lock cycle with DaemonState::lock and PyModules::lock.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-07-22 13:48:00 -04:00
luokexue
5ba8236e19 Update .organizationmap
The list of contributors to ZTE looks better if it is in alphabetical order.

Signed-off-by: Luo Kexue luo.kexue@zte.com.cn
2017-07-22 23:11:09 +08:00
songweibin
ba01a9462c doc: add rbd new trash cli in rbd.rst
Signed-off-by: songweibin <song.weibin@zte.com.cn>
2017-07-22 16:18:11 +08:00
Yan Jun
79168ecb6f osd: store op request source info
if we use command 'ceph daemon osd.# dump_historic_ops' to check the history
 tracked ops, the 'client_addr' is always dumped as '-' which result in clients
indistinguishable from each other.

"description": "osd_op(client.24123.0:328 3.36 3:6ec0d61b:::rbd_data.10472ae8944a.0000000000000289:head [write 4141056~4096 [fadvise_sequential]] snapc 0=[] ondisk+write+known_if_redirected e36)",
            "initiated_at": "2017-07-22 14:24:25.925633",
            "age": 4.893190,
            "duration": 0.272642,
            "type_data": {
                "flag_point": "commit sent; apply or cleanup",
                "client_info": {
                    "client": "client.24123",
                    "client_addr": "-",
                    "tid": 328
                },

It is because 'connection' was reset to nullptr in function '_unregistered'
before op has been inserted into history, so the 'get_source_addr' always return
empty entity addr.

Signed-off-by: Yan Jun <yan.jun8@zte.com.cn>
2017-07-22 15:29:39 +08:00
xie xingguo
fa0e314cde test: s/osd_objectstore_type/osd_objectstore/
Signed-off-by: xie xingguo <xie.xingguo@zte.com.cn>
2017-07-22 15:22:31 +08:00
Yan Jun
53c850541e common: dump field name of 'flag_point' and 'client_info'
we used to open "type_data" section as a array (that's open_array_section),
some key field names can not be dumped out. use open_object_section to dump them
as following:

before this change:
------------------
"description": "osd_op(client.4202.0:4958 3.28 3:163e8348:::rbd_data.10472ae8944a.0000000000000164:head [write 1691648~4096 [fadvise_sequential]] snapc 0=[] ondisk+write+known_if_redirected e24)",
            "initiated_at": "2017-07-22 14:22:13.469324",
            "age": 17.595664,
            "duration": 0.460406,
            "type_data": [
                "commit sent; apply or cleanup",
                {
                    "client": "client.4202",
                    "client_addr": "-",
                    "tid": 4958
                },
-----------------
after:
-----------------
"description": "osd_op(client.24123.0:328 3.36 3:6ec0d61b:::rbd_data.10472ae8944a.0000000000000289:head [write 4141056~4096 [fadvise_sequential]] snapc 0=[] ondisk+write+known_if_redirected e36)",
            "initiated_at": "2017-07-22 14:24:25.925633",
            "age": 4.893190,
            "duration": 0.272642,
            "type_data": {
                "flag_point": "commit sent; apply or cleanup",
                "client_info": {
                    "client": "client.24123",
                    "client_addr": "-",
                    "tid": 328
                },
----------------

Signed-off-by: Yan Jun <yan.jun8@zte.com.cn>
2017-07-22 14:27:55 +08:00
Sage Weil
4e6487cad4 Merge pull request #15991 from dillaman/wip-rbd-auth-profile
mon,osd: new rbd-based cephx cap profiles

Reviewed-by: Sage Weil <sage@redhat.com>
2017-07-21 22:38:42 -05:00
Sage Weil
a40d8e4a83 Merge pull request #16432 from liewegas/wip-mgr-daemonstate-lock
mgr: add per-DaemonState lock

Reviewed-by: John Spray <john.spray@redhat.com>
2017-07-21 22:36:26 -05:00
Sage Weil
0429acda45 Merge pull request #16460 from liewegas/wip-mgr-metadata
mon: add mgr metdata commands, and overall 'versions' command for all daemon versions

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-07-21 22:36:09 -05:00
Sage Weil
165f80a9e1 Merge pull request #16465 from xiexingguo/wip-shutup-noscrub-warn
mon: stop issuing not-[deep]-scrubbed warnings if disabled

Reviewed-by: Sage Weil <sage@redhat.com>
2017-07-21 22:35:43 -05:00
Sage Weil
31e9061815 Merge pull request #16477 from tchaikov/wip-health-mon
mon/HealthMonitor: trigger a proposal if stat updated

Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: xie xingguo <xie.xingguo@zte.com.cn>
2017-07-21 22:34:50 -05:00
Sage Weil
4d4b0b469e Merge pull request #16478 from xiexingguo/wip-drop-empty-message
mon/HealthMonitor: avoid sending unnecessary MMonHealthChecks to leader

Reviewed-by: Kefu Chai <kchai@redhat.com>
2017-07-21 22:34:32 -05:00
Sage Weil
bea20a4ce1 Merge pull request #16480 from jcsp/wip-rm-convert
common: remove config opt conversion utility

Reviewed-by: Ilya Dryomov <idryomov@redhat.com>
2017-07-21 22:34:08 -05:00
Sage Weil
c1704f2e6b mon: do not dereference empty mgr_commands
This may be empty before the initial quorum commits the static command
list.

Signed-off-by: Sage Weil <sage@redhat.com>
2017-07-21 23:32:05 -04:00
Sage Weil
2f272ab451 qa/workunits/cephtool/test.sh: add sudo for daemon compact
Signed-off-by: Sage Weil <sage@redhat.com>
2017-07-21 23:18:03 -04:00
Patrick Donnelly
1e5c8aef7b
Merge PR 16390 into master
* refs/remotes/upstream/pull/16390/head:
	mds: cleanup with unique_ptr
	mds: add to get version in the mds_commands
	mds: cleanup

Reviewed-by: Patrick Donnelly <pdonnell@redhat.com>
2017-07-21 13:23:10 -07:00
Patrick Donnelly
9506789ce1
Merge PR 16379 into master
* refs/remotes/upstream/pull/16379/head:
	qa: fix MDS_CLIENT_RECALL copy error

Reviewed-by: Zheng Yan <zyan@redhat.com>
2017-07-21 13:23:07 -07:00
Patrick Donnelly
9e936c70b2
Merge PR 16337 into master
* refs/remotes/upstream/pull/16337/head:
	mds: clear dirty rstat flag on non-auth inode when finishing scatter update

Reviewed-by: Zheng Yan <zyan@redhat.com>
2017-07-21 13:23:04 -07:00