Sage Weil
4a644e246e
msgr: don't initiate connect if we keepalive on unconnected peer
2009-11-17 16:30:39 -08:00
Sage Weil
6e88d27685
msgr: fix possible use-after-free by taking pipe lock during reap
...
This ensures that whoever called stop() (mark_down, in particular) finished
with the pipe (unlocked it) before we go and free it. Otherwise we might
call p->lock.Unlock() after the reaper deleted the Pipe, mucking up some
other memory.
I don't think this was actually triggerd, tho, since we would have seen
the assert(nlock == 0) in ~Mutex???
2009-11-17 16:30:36 -08:00
Sage Weil
302fc82f32
msgr: use common helper for reader/writer thread stop and reap queueing; fix locking
...
We weren't holding lock during ::close(sd)
2009-11-17 16:30:12 -08:00
Sage Weil
760165c64a
msgr: rename report_failures -> drop_msgs
...
Name was from legacy behavior
2009-11-17 14:03:36 -08:00
Sage Weil
318698b07e
msgr: small cleanups
2009-11-17 14:02:57 -08:00
Sage Weil
62ccae85ea
monc: small sanity check
2009-11-17 14:02:09 -08:00
Sage Weil
81fd4e159c
uclient: set mode on snapdir properly
2009-11-17 11:37:36 -08:00
Sage Weil
09938e891b
uclient: assert we have CAP_PIN
...
Not sure why that was conditional before??? Hmm!
2009-11-17 11:34:02 -08:00
Sage Weil
3e8e5f6e90
uclient: always retain CAP_PIN
2009-11-17 11:30:23 -08:00
Sage Weil
fcc234a8a0
monc: small cleanup
...
avoid unnec finish_hunting call on get_monmap_privately
2009-11-17 10:46:48 -08:00
Greg Farnum
cfd69fc48e
rados: Update stat calculation and display; change to wait_for_safe
2009-11-17 10:03:11 -08:00
Greg Farnum
a3cb25c955
qa: We want to stop on errors. Add new scripts.
2009-11-17 09:13:47 -08:00
Yehuda Sadeh
0669ca21f4
auth: reuse global_id when requesting tickets
2009-11-16 16:46:59 -08:00
Yehuda Sadeh
9731313896
vstart: fix overwrite conf
2009-11-16 16:46:59 -08:00
Sage Weil
fecee16d9d
mon: fix decode authmon stash on restart
2009-11-16 14:33:12 -08:00
Sage Weil
709be25a2b
mds: don't crap on !check_from, this doesn't happen with nolaggy
2009-11-16 14:25:30 -08:00
Sage Weil
8c6d2769b1
mon: move misplaced parens
2009-11-16 14:25:13 -08:00
Sage Weil
aa4e5b5752
todo
2009-11-16 13:35:45 -08:00
Sage Weil
f50a0517b3
msgr: get rid of harmless valgrind error
...
==7781== Source and destination overlap in memcpy(0x5B97EA8, 0x5B97EA8, 136)
2009-11-16 11:46:03 -08:00
Sage Weil
40edb9e253
mon: fail on write error
...
Don't rename badly written file. Also assert success, for now.
2009-11-16 11:45:25 -08:00
Sage Weil
37145dc154
monc: fix race/deadlock for get_monmap_privately
2009-11-13 16:38:52 -08:00
Sage Weil
a22418bb94
uclient: use Inode::auth_cap to find mds to write back snap cap metadata to
2009-11-13 16:27:26 -08:00
Sage Weil
e5ca40dd0c
uclient: keep MClientRequest local to send_request
2009-11-13 16:26:59 -08:00
Sage Weil
63f75e5608
mds: journal open_files based on is_any_caps_wanted(), not is_any_caps()
...
Actually we're a bit conservative in a few places since the wanted check
is a bit more expensive. We always do a full check in try_to_expire, so
much of the time we can do the quick check only.
2009-11-13 15:21:32 -08:00
Sage Weil
07369305f8
mds: don't rejournal files with caps that are unwanted
...
If they're unwanted, it's no biggie to fail to reconnect the cap. And
Locker::adjust_cap_wanted() already adjusts the open_file logseg lists in
this way, so let's just totally consistent.
2009-11-13 15:21:32 -08:00
Sage Weil
7e7a79e57b
mds: don't croak on open_files without caps
...
We can get a capless inode here if we replay an open file, the client
fails to reconnect it, but does REPLAY an open request (that adds it
to the logseg). AFAICS it's ok for the client to replay an open on a
file it doesn't have in it's cache anymore.
2009-11-13 15:21:32 -08:00
Greg Farnum
f3e746d5e6
rados: status printouts are now threaded
2009-11-13 11:56:42 -08:00
Sage Weil
ca342b8771
mkmonfs: cleanup monmap fn arg
2009-11-13 11:10:36 -08:00
Sage Weil
531f227f33
Makefile: include all the headers we need to build
2009-11-13 10:55:31 -08:00
Sage Weil
788ddaa319
config: accept list of possible conf file locations; include ./ceph.conf
...
This makes './ceph -w' work with './vstart.sh -x'.
2009-11-13 10:47:52 -08:00
Sage Weil
75a3c1ac24
vstart: -x to start up with cephx
2009-11-13 10:40:10 -08:00
Sage Weil
2acf8b61b0
monc: fix race/deadlock for get_monmap_privately
2009-11-13 10:40:02 -08:00
Sage Weil
9880cad39c
mds: don't try to send mdsmap to clients that need to reconnect
...
It won't work, clients must connect to us.
Also stop sending the map on recovery finish; they'll also (currently, at
least) get that from the monitor.
2009-11-13 10:09:31 -08:00
Sage Weil
2fb6f73618
mds: ignore/warn on late client reconnect attempts
...
In the future maybe we can do better (i.e. best effort attempt to reconnect
after the normal interval has passed)
2009-11-13 10:09:12 -08:00
Yehuda Sadeh
19087cf6cc
auth: don't try to read keys if cephx not supported
2009-11-13 09:34:51 -08:00
Yehuda Sadeh
96b5d97ab4
auth: use user admin as the default user
2009-11-12 16:54:37 -08:00
Greg Farnum
2303a33e77
Merge branch 'unstable' of ceph.newdream.net:/git/ceph into unstable
2009-11-12 16:45:05 -08:00
Greg Farnum
565448b5bd
qa: prepare for a hierarchical test script system
2009-11-12 16:44:35 -08:00
Sage Weil
27bc5741fc
vstart: some cleanup
2009-11-12 16:12:20 -08:00
Sage Weil
f8b0e54c74
conf: default to auth type 'none'
2009-11-12 15:29:36 -08:00
Sage Weil
1b5bb1fd6c
auth: fix compile errors, warnings
2009-11-12 15:29:26 -08:00
Sage Weil
2dcdd08339
auth: kill residual kernel/ files
2009-11-12 15:24:50 -08:00
Sage Weil
b5a11130be
Merge branch 'unstable' into auth
2009-11-12 15:23:24 -08:00
Sage Weil
a4cc231458
uclient: fewer msgs
2009-11-12 15:12:36 -08:00
Sage Weil
35e68924d1
todo
2009-11-12 15:12:18 -08:00
Sage Weil
e496deabf5
Merge branch 'unstable' of ssh://ceph.newdream.net/git/ceph into unstable
2009-11-12 14:55:26 -08:00
Sage Weil
35c3586bf4
put testing ceph.conf's in git
2009-11-12 14:57:11 -08:00
Sage Weil
9284e90aee
osd: print useful mode wait message
2009-11-12 14:57:11 -08:00
Sage Weil
c514691b27
filestore: skip sync_file_range if the full commit has started
...
Presumably btrfs commit_transaction is more effective than our per-file
flushing. And we don't want to fight each other.
2009-11-12 14:57:11 -08:00
Sage Weil
ad41e448b9
mds: recommit after commit if waiting for newer version
...
If there are waiters for a later version of the dir to hit
disk, then we need to recommit as soon as the prior commit
completes. We auth_pin on adding the first waiter, and do
not unpin until removing the last waiter, so this doesn't
break auth_pin rules.
Previously we could stall because we didn't finish the
waiter (on the later version) but also never started the
commit. Sometimes we would get lucky and someone else would
ask for a commit, but sometimes not. We would then see old
LogSegments that would never get fully expired.
2009-11-12 14:54:22 -08:00
