RepoMirrors/ceph
1
0
Fork 0
mirror of https://github.com/ceph/ceph synced 2025-02-15 14:58:01 +00:00
Code Issues Packages Projects Releases Wiki Activity
7,926 Commits 216 Branches 395 Tags 1.2 GiB
1b583a7ddf
Commit Graph

7926 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Sage Weil
03411a1565 mds: avoid propagating nestlock more frequently than mds_dirstat_min_interval 
Instead, add the lock to the update_scatterlock list, as usual.
 2009-08-24 14:01:14 -07:00
Sage Weil
9114e6fccb mds: kill SimpleLock::wait_shift 
It's a function of lock type.
 2009-08-24 12:45:48 -07:00
Sage Weil
c840ac9ad3 todo 2009-08-24 11:41:00 -07:00
Sage Weil
9d1bee70f1 uclient: warn about stat_precise struct changes 2009-08-24 11:33:51 -07:00
Sage Weil
10e644afcd libceph: no need for frag_info_t 2009-08-24 11:33:51 -07:00
Sage Weil
5d9d1ef61b msgr: add ms_die_on_bad_msg option, make assert(0) conditional 2009-08-24 11:33:51 -07:00
Greg Farnum
f6f126335d Hadoop: Style fixes. 2009-08-24 11:09:51 -07:00
Greg Farnum
cc7b36c0a0 Hadoop: Move the . and .. elimination 
from Java to C++ for less copying.
 2009-08-24 11:09:51 -07:00
Greg Farnum
4a6d22d86d uclient: _getdents should *always* return the number of bytes written 
at end of dir, not just assume it's zero!
 2009-08-24 11:09:51 -07:00
Greg Farnum
8d1f4d59d6 Hadoop: Use the new ceph_getdnames instead of getdir 
Hypertable: fix a nasty not-closing-the-dir bug
 2009-08-24 11:09:51 -07:00
Greg Farnum
93552e4199 libceph: Change the C++-only functions to be C-compatible. 
Hypertable: Change to use ceph_getdnames instead of getdir.

Hadoop: Minor updates to use the new C-style functions in libceph.
 2009-08-24 11:09:51 -07:00
Greg Farnum
af72edd717 Hadoop: Fix a nasty IO bug. 
CephInputStream.read() needs to return an int between 0 and 255,
but bytes are signed and casting preserves that. Adjust if needed.
 2009-08-24 11:09:51 -07:00
Greg Farnum
e9b00e09d2 Hadoop: Clean up ceph_delete code as it's redundant. 2009-08-24 11:09:51 -07:00
Greg Farnum
575158ab9f Hadoop: Fixed time modification in getFileStatus. 
No wonder I hadn't noticed an issue with statting!
 2009-08-24 11:09:51 -07:00
Greg Farnum
c5dd75eeb4 Hadoop: Add package.html; 
fix a NULL instead of null error (Java)
 2009-08-24 11:09:51 -07:00
Greg Farnum
bc372c7de7 Hadoop: Put guards up to prevent 
a nasty bug in case of unspecified preferences, and the unlikely case
that initialize is called twice.
 2009-08-24 11:09:51 -07:00
Sage Weil
1d7726ea9c upload tarball with 'stable' deb 2009-08-24 10:56:35 -07:00
Sage Weil
2b32297267 v0.13 2009-08-24 10:45:14 -07:00
Sage Weil
43dcbc2342 todo 2009-08-21 16:35:01 -07:00
Sage Weil
350190ccb4 kclient: cleanup msgpool pr_err 2009-08-21 14:36:54 -07:00
Sage Weil
d349d577bb kclient: update ctime on 'no-op' ->setattr(.., ATTR_CTIME) 
If kernel wants ctime updated but nothing else, pick a cap to
dirty, or set a no-op setattr req to the mds.

With this patch, we pass the full pjd fstest with mode ext3.
 2009-08-21 14:28:26 -07:00
Sage Weil
94b85eeff6 kclient tests: fix dir names 2009-08-21 13:49:46 -07:00
Sage Weil
411dcbea87 uclient: small _getdents cleanup 2009-08-21 13:44:13 -07:00
Sage Weil
2279e43546 uclient: let's do getdents() too, while we're at it 2009-08-21 13:27:31 -07:00
Sage Weil
3ccb16483a crush: put temp arrays on the stack 
40 ints on the stack.. not too much I hope?
 2009-08-21 13:20:42 -07:00
Sage Weil
206eddb15a uclient: remove readdirplus, readdirlite cruft 
Ancient!
 2009-08-21 13:14:50 -07:00
Sage Weil
8daf25b1a6 uclient: add getdnames() 2009-08-21 13:11:44 -07:00
Sage Weil
873a160388 kclient: better info in cap msg debug prints 2009-08-21 12:32:07 -07:00
Sage Weil
9397867446 mds: debug show cap issued/pending/wanted or issued/wanted, depending 2009-08-21 12:32:07 -07:00
Sage Weil
b180b6ce19 mds: clean out old revocation history on racing release 
If a release and revoke are racing, we need to make sure the MDS
behaves even though the revoke msg will be ignored.  For example:

mds                    client
1 pAs                  pAs
                       release(last_issue 1)
2 p                    p (*** ignored ***)
3 pAs (issue)          pAs
4 p                    p
5 pXs (issue)          pXs
release(last_issue 1)  *** (was completely) ignored ***

We fix this by including both seq AND last_issue in revoke_info.
The cap state for the above is thus:

mds cap state:
 last_issue = 5
 _pending = pXs
 _revokes = [ p seq=2 li=1, p seq=4 li=3 ]

and the MDS will trim any old revokes for which li <= its
last_issue.  For anything after that, the client will have seen a
new issue and will have stopped ignoring cap messages.
 2009-08-21 12:28:01 -07:00
Sage Weil
f9ec9d4a76 mount.ceph: block same signals as /sbin/mount 2009-08-21 11:00:33 -07:00
Sage Weil
5822c6a2c2 mds: remove mdstable ino cruft; version all ondisk encoding 2009-08-21 10:32:48 -07:00
Sage Weil
d149c7e7ec mds: do not save cap_push_seq on disk [disk format change] 
The cap_push_seq is purely associated with the current instance
of the session.  On MDS restart and client reconnect, it resets
to zero.  Don't save it.
 2009-08-21 10:28:41 -07:00
Sage Weil
28afd6958a kclient: make msgpools blocking or non-blocking 2009-08-21 09:32:51 -07:00
Sage Weil
f155bd8c53 kclient: stop using msgpools for incoming msgs 
Even though we may only dispatch N messages at a time (if we
have N cores), we may have many more connections, and be
reading the same message types in parallel from multiple
peers.  Msgpool_alloc blocks when teh pool is empty, which
will make the messenger thread hang, and stall forward.

Badness.

There is no easy solution here.
 2009-08-20 16:48:31 -07:00
Sage Weil
921260a17a kclient: be smarter about osd reply message refs 2009-08-20 16:48:31 -07:00
Sage Weil
a23e25f1d0 configure: check for expat for radosgw 2009-08-20 16:02:46 -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
0266f877ad kclient: don't keep reply ref in osd_request; just parse out the result 
This requires a few new flags: r_prepared_pages, so we only do that
once, and r_got_reply, so we only parse the result once.
 2009-08-20 14:45:51 -07:00
Sage Weil
7f320847fe kclient: add iozone, kernel tar+compile tests 2009-08-20 11:48:30 -07:00
Sage Weil
32a2b41e60 kclient: remove ancient crufty test code 2009-08-20 11:23:31 -07:00
Sage Weil
f8060443c1 kclient: use msgpool for client_caps 2009-08-19 16:15:37 -07:00
Sage Weil
065b1fdc65 kclient: clean up msgr debug output a bit 2009-08-19 13:52:19 -07:00
Sage Weil
d142cf314f kclient: fix data skip arithmetic 2009-08-19 13:52:18 -07:00
Sage Weil
11376ca2e7 kclient: drop middle, data in ceph_msg_put 
And not msgpool reclaim.

This fixes bug where a reused message has nr_pages != 0 and
gets confused.
 2009-08-19 13:52:17 -07:00
Sage Weil
a849be3242 auth: auth scheme notes 2009-08-19 10:18:29 -07:00
Sage Weil
f60482450f Merge branch 'unstable' of ssh://ceph.newdream.net/git/ceph into unstable 2009-08-19 10:15:34 -07:00
Sage Weil
8d223df7db kclient: requeue cap if cap send fails 2009-08-19 10:14:49 -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