Commit Graph

622 Commits

Author SHA1 Message Date
Sage Weil
8e55d56c01 todo 2009-10-06 14:00:54 -07:00
Sage Weil
854d6645da todo 2009-10-02 15:29:44 -07:00
Sage Weil
5b41c13f0d todo 2009-10-01 13:28:30 -07:00
Sage Weil
6350e849ca todo 2009-10-01 12:57:39 -07:00
Sage Weil
bce85f1cc5 kclient: debugfs fix; better osdmap debug output 2009-09-30 14:10:46 -07:00
Sage Weil
15d45e76bb kclient: include new files in patch set 2009-09-30 12:53:49 -07:00
Sage Weil
21df30f61f document data structs; kill ttl_ms in caps msg 2009-09-30 12:34:30 -07:00
Sage Weil
25bba5df97 mds: kill old rdcaps_ttl business
This was from when caps independently timed out.
2009-09-30 12:24:31 -07:00
Sage Weil
4874b1901b kclient: try kmalloc, then vmalloc
"A still-lame-but-less-lame option here would be to attempt the kmalloc
(with __GFP_NOWARN) and if it failed, fall back to vmalloc." says akpm
2009-09-30 11:43:31 -07:00
Sage Weil
59fc0e27d5 kclient: more fixes from akpm's review 2009-09-30 11:01:37 -07:00
Sage Weil
c064c4456d uninline some header types 2009-09-30 10:24:28 -07:00
Sage Weil
bc919199ca mdsmap: use sized int type 2009-09-29 14:53:15 -07:00
Sage Weil
4dbb6fdfdb todo 2009-09-29 14:53:15 -07:00
Sage Weil
10e4b6ea33 uclient: get_local_osd() to find an osd with the given ip 2009-09-28 13:47:22 -07:00
Sage Weil
281acde0c1 kclient: user.ceph.layout readonly for now 2009-09-28 10:56:05 -07:00
Sage Weil
6c8d6a530f todo 2009-09-23 10:11:56 -07:00
Sage Weil
d9f94b2de4 todo 2009-09-22 14:18:50 -07:00
Sage Weil
13e6f345cf todo 2009-09-22 10:00:31 -07:00
Sage Weil
cf3192dd39 todo 2009-09-21 16:52:40 -07:00
Sage Weil
c2c9ecd1dd mon: fix up subscribe infrastructure to use msgr connection for sessions
Session state is now cleaned up when connections drop.  More
efficient in-memory structures.
2009-09-21 14:30:19 -07:00
Sage Weil
dbfb0f0deb kclient: add GET_DATALOC ioctl to get object location, osd addr 2009-09-18 12:45:48 -07:00
Sage Weil
0f892e34be todo 2009-09-17 13:47:35 -07:00
Sage Weil
c7b904e213 todo 2009-09-17 13:18:08 -07:00
Sage Weil
131fa21c10 kclient: move caps stats into client debugfs dir; fix error checking 2009-09-17 12:04:35 -07:00
Sage Weil
cca45f3c7b uclient: do not do locking for 'atomic' writes
They aren't tested and aren't safe (client may fail; there's no cleanup).
2009-09-15 13:14:01 -07:00
Sage Weil
4fddaad08d debian: fix radosgw package 2009-09-10 15:10:29 -07:00
Sage Weil
d20510956e kclient: fix format strings for entity num 2009-09-10 14:32:05 -07:00
Sage Weil
122c612156 kclient: fix checkpatch label warnings 2009-09-10 11:23:33 -07:00
Sage Weil
5f3b046982 kclient: allow sent msgs to be revoked from a connection
The simplest solution seemed to be to replace the spinlock with a
mutex and carry it for the duration of try_write() (including the
NOWAIT calls to kernel_sendmsg, etc).  If there is contention the
overhead will be a bit higher, but the code is much simpler.
2009-09-09 16:46:42 -07:00
Sage Weil
69f84d1b77 todo 2009-09-08 14:48:27 -07:00
Sage Weil
900215b9cb todo 2009-09-08 13:46:42 -07:00
Sage Weil
5bc0dea9ff msgr todos 2009-09-03 10:39:07 -07:00
Sage Weil
4330037a69 msgr: include my addr AND socket addr during connect handshake
This allows connecting side to detect their ip address.
2009-09-02 16:10:05 -07:00
Sage Weil
a2c869c3ce mds: use client_t throughout 2009-09-02 12:53:47 -07:00
Sage Weil
5901023c16 kclient: osd restart fixes 2009-09-01 16:07:55 -07:00
Sage Weil
66efd409db kclient: fail con->ops->get() if ref count is zero, and avoid requeueing work
This fixes a bug where a socket status change races with con_destroy:

ref 1 -> 0
	con_destroy()
					sock state change, requeue
		destroy socket
		kfree(con)
					...do work and crash...
2009-08-27 09:14:15 -07:00
Sage Weil
207880e9be kclient: factor out connection creation; add con_send, con_keepalive
Mmm, much cleaner so far!
2009-08-26 13:09:09 -07:00
Sage Weil
0cfc905e15 msgr: server flag to avoid connecting to peer
If we have a message and no socket and server=true, just queue it
up.  When they connect to us they'll get it.
2009-08-26 13:09:09 -07:00
Sage Weil
1e5e8d8885 mds: update dir parent pointer on rename
Maintain per-segment list and inode flag for this purpose.
Flush out changes on log segment trim.

Also remove from renamed list if we commit first frag (which also
writes a parent pointer each time).
2009-08-25 14:56:23 -07:00
Sage Weil
c840ac9ad3 todo 2009-08-24 11:41:00 -07:00
Sage Weil
43dcbc2342 todo 2009-08-21 16:35:01 -07:00
Sage Weil
10940c2c81 Makefile: add missing MOSDPGTemp.h 2009-08-20 15:52:24 -07:00
Sage Weil
c55f2e9812 mds: clean up loner set/drop logic
Track what loner we _want_ separately from what loner we have, so
that locks can move toward the right state.

Centralize loner set/drop call points.

Move out of EXCL in simple_eval if our target loner is -1(none).
2009-08-20 14:48:10 -07:00
Sage Weil
6d8144c04b kclient: don't reuse message for revoke_ack
This is error prone (multiple paths for sending caps) and screws
up the msg pools.
2009-08-19 09:53:17 -07:00
Sage Weil
dc31490da6 todo 2009-08-14 11:57:19 -07:00
Sage Weil
35bac67a60 mds: rewrite Capability issue/revoke tracking
Use a list.  Track everything.  Be smart about trimming the list.
2009-08-14 10:56:39 -07:00
Sage Weil
58483919cf todo 2009-08-13 16:51:06 -07:00
Sage Weil
2c2ee44e3f todo 2009-08-13 15:33:51 -07:00
Sage Weil
0d45c57cad Makefile: include stop.sh in tarball 2009-08-11 16:29:01 -07:00
Sage Weil
a5b60b5157 todo 2009-08-11 16:22:11 -07:00