Sage Weil
2368a22365
osd: destroy messengers on shutdown
2009-06-10 14:04:52 -07:00
Sage Weil
8ac2665b71
monclient: avoid hang during startup
2009-06-10 14:04:35 -07:00
Sage Weil
93984cfca7
osd: fix huge memory leak (MOSDOp)
2009-06-10 13:44:52 -07:00
Sage Weil
73ac6f31a8
radostool: select snaps, rm objects
2009-06-10 13:44:52 -07:00
Yehuda Sadeh
e4fa8f5897
rados: get rid of wrcall
2009-06-10 12:53:35 -07:00
Yehuda Sadeh
c8b162fb94
rados: rename rdcall to call
2009-06-10 12:16:47 -07:00
Yehuda Sadeh
a1b134ad2c
class: add cls_setxattr
2009-06-10 11:18:17 -07:00
Yehuda Sadeh
43439e2e53
class: set method flags on registration
2009-06-10 10:50:25 -07:00
Yehuda Sadeh
659b20ea6a
osd: don't call may_read/may_write prematurely (revert previous)
2009-06-10 10:26:42 -07:00
Yehuda Sadeh
91f0af3ec1
osd: update rwm flags for incoming ops
2009-06-10 10:04:11 -07:00
Sage Weil
0a3ae3608a
radostool: clean up bench args, usage
2009-06-09 21:24:29 -07:00
Sage Weil
3a3bd4f6ad
osd: include size in object_info_t
...
The main advantage here, aside from slightly decluttering the
code, is to avoid a stat() in get_object_context. We just
getxattr and that will implicitly tell us if the object
exists. (If the file exists but the OI_ATTR xattr doesn't,
the object is busted anyway.)
2009-06-09 21:24:26 -07:00
Sage Weil
8ac00cb15c
osd: include ondisk magic in encoding for key ondisk state structs
...
This will eventually allow us to incrementally migrate to
updated ondisk structures by making the decoding conditional
(dependent on the encoded version). At least in some cases.
2009-06-09 20:36:44 -07:00
Sage Weil
aed778a64a
rados: remove testradoscio[pp] targets
2009-06-09 20:36:44 -07:00
Sage Weil
5052dc8e3a
rados: add helper to header
2009-06-09 20:36:44 -07:00
Sage Weil
931b0d7972
buffer: remove unused field
2009-06-09 20:36:43 -07:00
Sage Weil
8502be2188
osd: fix up OSDOp printing
2009-06-09 20:36:43 -07:00
Greg Farnum
109ed973e7
radostool now includes benchmarking functionality from testradosciopp.cc
2009-06-09 17:02:00 -07:00
Sage Weil
0660c2ca6e
osd: rev osd protocol
2009-06-09 16:48:41 -07:00
Sage Weil
70efc45883
kclient: include tid in osdc debugfs dump
2009-06-09 16:34:48 -07:00
Yehuda Sadeh
7a624a4a8a
osd: hold op data on the specific op
2009-06-09 16:33:30 -07:00
Sage Weil
82e14ae4a1
msgr: calculate data crc after encode_payload, in worker thread
...
Faster (bc crc calc is in msgr worker thread). Also allows
encode_payload() to manipulate data.
2009-06-09 16:02:36 -07:00
Greg Farnum
5317718464
Removed time_t -> int conversion warnings.
2009-06-09 16:01:58 -07:00
Sage Weil
01d91acbce
mds: fix dir frag split replay
...
is_complete() may not be set, but split is still allowed.
2009-06-09 15:22:55 -07:00
Sage Weil
399f22cb18
osd: don't clobber pg_stats.reported when repairing stats
2009-06-09 15:22:18 -07:00
Sage Weil
ef5f8f61e1
mds: do not fragment dirs if mds_bal_fragment_interval <= 0
2009-06-09 15:07:24 -07:00
Sage Weil
debd502126
radostool: cleanup
2009-06-09 14:58:53 -07:00
greg
60e3862709
Hurray, this file works properly as a standalone.
...
It takes concurrent writes to maintain, minimum runtime, size of the writes, and whether or not to check for consistency. Reports total runtime, writes made, bandwidth, average & max latency, and time waiting for Rados.
2009-06-09 14:56:08 -07:00
greg
6f47790ba0
Added initial testradosciopp.cc.
2009-06-09 14:56:08 -07:00
greg
5792669a15
A properly working testradoscio.c. For the moment it just takes a number of writes to make and does that number in a loop.
2009-06-09 14:56:08 -07:00
Sage Weil
ae4d91ccd8
osd: simplify op read/write flag setup
2009-06-09 14:49:43 -07:00
Sage Weil
ca27616e6c
osd: kick pg on active set change
2009-06-09 14:43:33 -07:00
Sage Weil
efdf8947ad
common: kill obsolete common/ThreadPool.h
2009-06-09 14:43:33 -07:00
Sage Weil
abb9fa4ce1
osd: fix osd shutdown vs scrub work queue deadlock
...
Pause new disk_tp items, kick all (possibly) blocked pgs, and _then_
stop the disk_tp. And make scrub() bail out if the osd is stopping.
2009-06-09 14:43:33 -07:00
Sage Weil
799723acb0
mon: fix 'scrub' -> 'repair' in command reply msg
2009-06-09 14:43:33 -07:00
Sage Weil
1a35911c79
osd: counnt errors fixed during repair, and indicate scrub vs repair
2009-06-09 14:43:33 -07:00
Sage Weil
daeed421ac
osd: fix pg stats on repair
2009-06-09 14:43:33 -07:00
Sage Weil
3c91cb5a48
osd: replica should not check missing map on push
...
That is, we should let the primary push us objects even if we don't
think they're missing. This is necessary for pg repair to work.
The old check is useless, since we already restart peering when the
pg changes, and ignore pre-change sub_op messages.
2009-06-09 14:43:33 -07:00
Sage Weil
7b506643ea
osd: improved logging for scrub
2009-06-09 14:43:33 -07:00
Sage Weil
7f01a4b26c
osd: fix repair_object attr name
...
Didn't get switched from 'oi' to '_' way back when.
2009-06-09 14:43:33 -07:00
Sage Weil
9f8fc96140
osd: fix pg parsing
2009-06-09 14:43:32 -07:00
Yehuda Sadeh
85a65e5fc6
osd: read/write op flags revision (for class methods)
2009-06-09 13:26:06 -07:00
Yehuda Sadeh
bdfbc610b5
class: class acl stuff preliminary stuff
2009-06-09 13:26:06 -07:00
Sage Weil
cb6e7007e6
mon: don't dup entires in log.err
2009-06-09 11:08:11 -07:00
Sage Weil
916b1998ce
osd: fix backlog ordering part deux
...
Remove the old entry from the map!
2009-06-09 10:57:59 -07:00
Sage Weil
24dd3ee903
osd: repair out of order logs in read_log
...
Works around bug fixed in f1ff93.
2009-06-09 10:36:03 -07:00
Sage Weil
f1ff934f94
osd: fix order of prior_version backlog entries
...
They need to end up in the log in the proper order. So make 2 passes,
one to generate the entries, and a second to prepend them to the log.
2009-06-08 17:36:43 -07:00
Sage Weil
61995ee7a6
osd: cleanup
2009-06-08 15:59:24 -07:00
Sage Weil
e16ed9b17b
mon: handle failures during CLIENTREPLAY state
2009-06-08 15:59:24 -07:00
Sage Weil
7685e7f823
radostool: 'ls' fixes
2009-06-08 14:11:54 -07:00
