RepoMirrors/ceph
1
0
Fork 0
mirror of https://github.com/ceph/ceph synced 2025-03-11 02:39:05 +00:00
Code Issues Packages Projects Releases Wiki Activity
70,881 Commits 219 Branches 396 Tags 1.4 GiB
37e9a874af
Commit Graph

70881 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Sage Weil
37e9a874af Merge pull request #13968 from dzafman/wip-15912-followon 
osd,mon: misc full fixes and cleanups

Reviewed-by: Sage Weil <sage@redhat.com>
Reviewed-by: John Spray <john.spray@redhat.com>
Reviewed-by: Kefu Chai <kchai@redhat.com>
 2017-04-17 16:42:13 -05:00
Yuri Weinstein
287f94d0c0 Merge pull request #14440 from liewegas/wip-status-flags 
osd/OSDMap: hide require_*_osd and sortbitwise flags


Reviewed-by: Josh Durgin <jdurgin@redhat.com>
 2017-04-17 13:34:37 -07:00
David Zafman
3becdd3138 test: Test health check output for full ratios 
Test out of order ratios summary and details
Test various full osd conditions summary and details

Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 13:02:57 -07:00
Matt Benjamin
b01bf489ad Merge pull request #14561 from linuxbox2/wip-rgw-reread-dir 
rgw_file: fix readdir after dirent-change
 2017-04-17 14:58:35 -04:00
Casey Bodley
7e19db2c2f Merge pull request #14466 from fangyuxiangGL/bi 
rgw: bucket stats display bucket index type

Reviewed-by: liuchang0812 <liuchang0812@gmail.com>
Reviewed-by: Casey Bodley <cbodley@redhat.com>
 2017-04-17 11:04:02 -04:00
David Zafman
2522307865 mon, osd: Add detailed full information for now in the mon 
Show ceph health doc output in the correct order

Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 08:02:50 -07:00
David Zafman
e4cf10d3d8 mon: Issue warning or error if a full ratio out of order 
The full ratios should be in this order: nearfull, backfillfull, full, failsafe full

Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 08:02:50 -07:00
David Zafman
c83f11de00 mon: Always fix-up full ratios when specified incorrectly in config 
Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 08:02:50 -07:00
David Zafman
afd739bed6 mon: Use currently configure full ratio to determine available space 
This is a bug that would not adjust available space based on the
currently configured full ratio, but rather the mon_osd_full_ratio
default initial value.

Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 08:00:24 -07:00
David Zafman
1fafec2175 osd: check_full_status() remove bogus comment and use equivalent computation 
We actually compute kb_used as the kb - kb_avail.  We don't have the
statfs() system call issue of non-privileged f_bavail vs f_bfree.  It
was assumed that used was really like (blocks - f_bfree).  It is not.

Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 08:00:24 -07:00
David Zafman
84088568b5 osd: Check whether any OSD is full before starting recovery 
Add event RecoveryTooFull to move to NotRecovering state

Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 08:00:24 -07:00
David Zafman
27e14504f6 osd: Add PG state and flag for too full for recovery 
New state machine state NotRecovering
New PG state PG_STATE_RECOVERY_TOOFULL

Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 08:00:24 -07:00
David Zafman
c7e8dcad34 osd: Add check_osdmap_full() to check for shard OSD fullness 
Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 08:00:24 -07:00
David Zafman
94e253ce37 osd: Rename backfill_request_* to recovery_request_* 
To be used by both recovery and backfill

Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 08:00:24 -07:00
David Zafman
1711ccdec7 osd: Check failsafe full and crash on push/pull 
Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 08:00:24 -07:00
David Zafman
1e2fde1012 osd: Revamp injectfull op to support all full states 
Use check_* for injectable full checks
Use is_* to just test simple cur_state

Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 08:00:24 -07:00
David Zafman
a5731076ad osd: Handle backfillfull_ratio just like nearfull and full 
Add BACKFILLFULL as a local OSD cur_state
Notify monitor of this new fullness state

Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 08:00:24 -07:00
David Zafman
0264bbddb7 osd: For testing full disks add injectfull socket command 
Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 07:58:30 -07:00
David Zafman
9dd6952999 common: Bump ratio for backfillfull from 85% to 90% 
Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 07:58:30 -07:00
David Zafman
79a4ac41c5 common: Remove unused config option osd_recovery_threads 
Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 07:58:30 -07:00
David Zafman
79124330c7 osd: too_full_for_backfill() returns ostream for reason 
Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 07:58:30 -07:00
David Zafman
5baf7abfa3 osd: Fail-safe full is a hard stop even for mds 
We can't allow OSD to become non-startable even if mds
could be writing as part of file removals.

Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 07:58:30 -07:00
David Zafman
e927cd2cf2 test: Fix intended test flow and restore nearfull-ratio 
This is inconsequential but seems to have always been wrong since original
commit 6cafb0e3e0

Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 07:58:30 -07:00
David Zafman
811f89a682 test: Switch from pg to osd for set-*-ratio commands 
Testing of 6422e0a220

Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 07:58:30 -07:00
David Zafman
d024ab0504 osd: Remove unused argument to clear_queued_recovery 
Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 07:58:30 -07:00
David Zafman
6c930f7e66 osd: Increase osd_backfill_retry_interval to 30 seconds 
Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 07:58:30 -07:00
David Zafman
3a66f1fbf6 ceph-objectstore-tool: cleanup comment 
Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 07:58:30 -07:00
David Zafman
1e7d227bfc osd: Fix log message 
Signed-off-by: David Zafman <dzafman@redhat.com>
 2017-04-17 07:58:30 -07:00
Casey Bodley
ac77cadc8d Merge pull request #12536 from ilc/cls_rgw_minor_refactor 
cls/rgw: Clean up the "magic string" usage in the cls layer for RGW.

Reviewed-by: Casey Bodley <cbodley@redhat.com>
Reviewed-by: Adam C. Emerson <aemerson@redhat.com>
 2017-04-17 09:40:47 -04:00
John Spray
1247119ca8 Merge pull request #14046 from songbaisen/m7 
mgr: optimization some judgment and adjust the debug remove value in register_new_pgs

Reviewed-by: John Spray <john.spray@redhat.com>
 2017-04-17 11:23:22 +01:00
songbaisen
3da07ea0a5 mgr: optimization some judgment and adjust the debug remove value in register_new_pgs 
Signed-off-by: song baisen <song.baisen@zte.com.cn>
 2017-04-17 06:53:26 +00:00
John Spray
4314e3d1c5 Merge pull request #14267 from yunfeiguan/master 
cephfs: fix mount point break off problem after mds switch occured 

Reviewed-by: Yan, Zheng <zyan@redhat.com>
 2017-04-15 19:46:53 +01:00
John Spray
c65e4fbf8e Merge pull request #13587 from yanghonggang/master 
cephfs: fix write_buf's _len overflow problem

Reviewed-by: John Spray <john.spray@redhat.com>
 2017-04-15 19:45:54 +01:00
John Spray
177699981e Merge pull request #13522 from singler/wip-reconnect-client 
client/Client.cc: after reset session from MDS - reconnect

Reviewed-by: Yan, Zheng <zyan@redhat.com>
 2017-04-15 19:44:53 +01:00
John Spray
0cb0c494d9 Merge pull request #13466 from jcsp/wip-cephfs-ecoverwrites 
Add test for is_hacky_ecoverwrites in cephfs pool checks

Reviewed-by: Kefu Chai <kchai@redhat.com>
 2017-04-15 19:44:00 +01:00
John Spray
f412345590 Merge pull request #14207 from jcsp/wip-19401 
mds: ignore ENOENT on writing backtrace

Reviewed-by: Yan, Zheng <zyan@redhat.com>
 2017-04-15 19:40:59 +01:00
Matt Benjamin
e0f80266ec rgw_file: fix readdir after dirent-change 
Also, fixes link count computation off-by-one, update of state.nlink
after computation, link computation reset at start, and a time print
in debug log.

Fixes: http://tracker.ceph.com/issues/19634

Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>

link count

Signed-off-by: Matt Benjamin <mbenjamin@redhat.com>
 2017-04-15 10:05:24 -04:00
John Spray
39bd183349 Merge pull request #13579 from ukernel/wip-19022 
client: drop cap snaps when auth mds session gets closed

Reviewed-by: John Spray <john.spray@redhat.com>
 2017-04-15 12:15:04 +01:00
Sage Weil
d82ac99d70 Merge pull request #14490 from liewegas/wip-no-luminous 
mon: add mon_debug_no_require_luminous

Reviewed-by: Kefu Chai <kchai@redhat.com>
 2017-04-14 20:05:45 -05:00
Matt Benjamin
260de938f0 Merge pull request #14553 from guihecheng/rgw_file-revert 
rgw_file: fix up potential race condition

verified f23
 2017-04-14 14:30:38 -04:00
Kefu Chai
43a2f3ed33 Merge pull request #14546 from dachary/wip-19628-directory-boot 
ceph-disk: enable directory backed OSD at boot time

Reviewed-by: Kefu Chai <kchai@redhat.com>
 2017-04-15 00:06:50 +08:00
Sage Weil
8c6820a433 qa/suites/rados/thrash: do not thrash cluster full without luminous flag 
The interface changes with the luminous flag; this will fail.

Signed-off-by: Sage Weil <sage@redhat.com>
 2017-04-14 12:04:47 -04:00
Sage Weil
4857f51e68 qa/suites/rados: expand other collections with no-require-luminous 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-04-14 11:45:05 -04:00
Kefu Chai
a29be2ad62 Merge pull request #14551 from tchaikov/wip-19584 
mon/OSDMonitor: transit creating_pgs from pgmap when upgrading

Reviewed-by: Sage Weil <sage@redhat.com>
 2017-04-14 22:42:34 +08:00
Sage Weil
2a4e431599 Merge pull request #14435 from liewegas/wip-bluestore-deferred 
os/bluestore: tune deferred_batch_ops separately for hdd and ssd

Reviewed-by: Igor Fedotov <ifedotov@mirantis.com>
 2017-04-14 08:35:50 -05:00
Sage Weil
08d3fa490f Merge pull request #14489 from liewegas/wip-bluestore-fix-flush 
os/bluestore: fix narrow osr->flush() race
 2017-04-14 08:35:35 -05:00
Sage Weil
16d3a524d8 mon/OSDMonitor: require_luminous_osds for pg-remap[-items] 
Signed-off-by: Sage Weil <sage@redhat.com>
 2017-04-14 09:21:17 -04:00
Loic Dachary
f425a127b7 ceph-disk: enable directory backed OSD at boot time 
539385b143
introduced a regression preventing directory backed OSD from starting at
boot time.

For device backed OSD the boot sequence starts with ceph-disk@.service
and proceeds to

    systemctl enable --runtime ceph-osd@.service

where the --runtime ensure ceph-osd@12 is removed when the machine
reboots so that it does not compete with the ceph-disk@/dev/sdb1 unit at
boot time.

However directory backed OSD solely rely on the ceph-osd@.service unit
to start at boot time and will therefore fail to boot.

The --runtime flag is selectively set for device backed OSD only.

Fixes: http://tracker.ceph.com/issues/19628

Signed-off-by: Loic Dachary <loic@dachary.org>
 2017-04-14 12:14:18 +02:00
Gui Hecheng
a32add15b8 rgw_file: add locks to fix potential race 
Signed-off-by: Gui Hecheng <guihecheng@cmss.chinamobile.com>
 2017-04-14 17:53:04 +08:00
John Spray
d51c027311 Merge pull request #13859 from jcsp/wip-19204 
mds: shut down finisher before objecter

Reviewed-by: Yan, Zheng <zyan@redhat.com>
 2017-04-14 10:50:14 +01:00