Yehuda Sadeh
1b6450b2c3
kclient: remove /sys/fs/ceph when doing cleanup
2009-03-04 10:34:27 -08:00
Sage Weil
2727657b6c
mds: logger typo
2009-03-03 16:59:17 -08:00
Sage Weil
97ec829aac
cosd: fixes
2009-03-03 16:40:47 -08:00
Sage Weil
7762ea920c
mkcephfs: bit more verbose
2009-03-03 16:40:47 -08:00
Sage Weil
4e04c53568
configure: remove some silly checks
2009-03-03 16:45:58 -08:00
Sage Weil
98e8d362f9
makefile: include make_version in dist tarball
2009-03-03 16:45:20 -08:00
Sage Weil
1fa26111f6
todo
2009-03-03 16:35:53 -08:00
Sage Weil
5e61967a6e
mds: fix logger key id; call validate() for all logtypes
2009-03-03 16:35:53 -08:00
Sage Weil
af870763f3
config: no more .ceph_monmap; parse cluster.conf for mon addrs
...
Look for cluster.conf in /etc/ceph/cluster.conf, then
./cluster.conf. Make vstart.sh generate a cluster.conf.
No more .ceph_monmap default.. only look for monmap file is
specified on command line.
2009-03-03 16:35:53 -08:00
Sage Weil
a273072d30
todo: cas notes
2009-03-03 16:35:53 -08:00
Sage Weil
6a53bfc13c
mds: small cleanups
2009-03-03 16:35:52 -08:00
Sage Weil
ea10a67267
mds: fix file_eval
...
Be more aggressive about shifting filelock state to something more
useful. Mainly, do NOT stop everything if there are wrlocks! This
causes all kinds of annoying stalls waiting for the log to flush.
2009-03-03 16:35:52 -08:00
Yehuda Sadeh
e66072fa46
ceph_ver: write log message whenever a new log is opened
2009-03-03 16:07:10 -08:00
Yehuda Sadeh
e52e455804
ceph_ver: add ceph_ver.h to list of .h files
2009-03-03 16:03:57 -08:00
Yehuda Sadeh
ed907427e1
kclient: write ceph git version when module loads
2009-03-03 16:01:48 -08:00
Yehuda Sadeh
6733493b55
ceph_ver: add unconditional rule
2009-03-03 15:46:06 -08:00
Yehuda Sadeh
b93190660a
ceph_ver: uncomment important line
2009-03-03 15:37:02 -08:00
Yehuda Sadeh
c88c4da5d2
ceph_ver: show the git sha1 version in the logs
2009-03-03 15:32:20 -08:00
Yehuda Sadeh
06edba73cb
Merge branch 'unstable' of ssh://ceph.newdream.net/git/ceph into unstable
2009-03-03 15:23:12 -08:00
Yehuda Sadeh
c2ba9b2030
kclient: wakeup wq waiting on caps on write
2009-03-03 14:35:56 -08:00
Sage Weil
0d4374764d
kclient: some fmode cleanup
...
BUG instead of WARN. Rename field.
2009-03-03 14:27:15 -08:00
Sage Weil
4bab554b90
kclient: fix fmode ref count buf
...
We weren't taking an fmode reference on the root dir. This would
cause a WARN if our root caps were allowed to expire and we had
to ropen them. Duplicated wonky r_fmode conditional from later
fill_inode() call in fill_trace().
2009-03-03 14:26:28 -08:00
Sage Weil
93ef8a3667
kclient: wake up i_cap_wq in ceph_add_cap
...
We could a thread blocked on caps when we are issued a new cap in
another thread. They need to be woken up.
2009-03-03 12:28:11 -08:00
Sage Weil
e508fb2ee3
kclient: check d_inode before clearing its COMPLETE flag.
2009-03-03 12:27:31 -08:00
Sage Weil
6f48094e5c
mds: fix try_remove_unlinked_dn
...
Just because a dir is_complete() doesn't mean that we can safely
do away with a null dentry. What matters is if the dentry existed
before the last commit.. and having everything in our cache isn't
relevant at all there. If we are complete, a fetch() would be a
no-op, but since we can clear complete at any time, that's not
helpful.
2009-03-03 12:25:58 -08:00
Sage Weil
ff2a6f6b96
kclient: do not return new null dentry from lookup
...
The VFS doesn't appear to like it when lookup returns a replacement
dentry that is null... it BUGs on unmount with a ref count that is
still 1.
So, if we get ENOENT and no trace, return ERR_PTR(-ENOENT) instead.
2009-03-03 11:43:47 -08:00
Sage Weil
fcbbe9ab2a
cfuse: fix startup
2009-03-03 11:33:42 -08:00
Yehuda Sadeh
4a5dbb700a
conf: some cleanup
2009-03-03 10:43:35 -08:00
Yehuda Sadeh
b0b561603a
conf: some more fixes (again)
2009-03-03 10:41:21 -08:00
Yehuda Sadeh
4570354050
conf: fix typo
2009-03-03 10:40:15 -08:00
Sage Weil
212cb785f8
mds: clean up auth_pin nesting code a bit
2009-03-03 10:42:21 -08:00
Sage Weil
ea1962ef1e
mkcephfs: take optional crush map on command line
2009-03-03 10:42:21 -08:00
Yehuda Sadeh
95bfdd784d
conf: lockdep again
2009-03-03 10:37:29 -08:00
Yehuda Sadeh
88334c2646
conf: fix some other errors
2009-03-03 10:17:13 -08:00
Yehuda Sadeh
ad13632d5b
conf: the return of --lockdep, other fixes
2009-03-03 10:14:17 -08:00
Yehuda Sadeh
6fdcf7421f
config: some cleanup
2009-03-03 10:06:42 -08:00
Sage Weil
2f3a19e9fd
debian: fix dependencies
2009-03-03 09:32:23 -08:00
Yehuda Sadeh
d9f8db2572
conf: fix erronous scoping
2009-03-02 17:19:47 -08:00
Yehuda Sadeh
0fc4a23e3f
conf: moved options to the preparse section
2009-03-02 17:12:35 -08:00
Yehuda Sadeh
13b1141e81
Merge branch 'unstable' of ssh://ceph.newdream.net/git/ceph into unstable
2009-03-02 17:05:23 -08:00
Yehuda Sadeh
3a116a0681
conf: added old remarks
2009-03-02 17:05:08 -08:00
Sage Weil
c99ecb9365
config: remove 'debug_after' cruft
2009-03-02 16:56:45 -08:00
Yehuda Sadeh
a7ee341c63
conf: removed some more old stuff
2009-03-02 16:40:11 -08:00
Yehuda Sadeh
8a6138542a
conf: remove old stuff
2009-03-02 16:37:56 -08:00
Sage Weil
69b7578567
kclient: revamp async truncation
...
We can learn about truncations from an mds reply or from a TRUNC
cap message. Set up the async truncation in the generic
ceph_fill_file_size helper. Remove old crufty async trunc code
(that assumed TRUNC would come before an ltruncate reply).
2009-03-02 16:38:15 -08:00
Sage Weil
e518e4a146
mds: flush log on wrlock_start if lock isn't stable
2009-03-02 16:38:15 -08:00
Sage Weil
fc72a5a21f
mds: flush mdlog on max_size increase
...
Presumably a writer is waiting for it.
2009-03-02 16:38:15 -08:00
Sage Weil
648349d195
mds: only issue cap TRUNC if truncating smaller
2009-03-02 16:38:15 -08:00
Sage Weil
97e2071333
filestore: more helpful error message with xattrs don't work
2009-03-02 16:38:15 -08:00
Sage Weil
2ac56cc0ec
sample.cluster.conf: fix mon path/osd dev option names
2009-03-02 16:38:15 -08:00
