]>
git.apps.os.sepia.ceph.com Git - ceph.git/log
Sage Weil [Thu, 8 Apr 2010 15:49:27 +0000 (08:49 -0700)]
buffer: move to ceph namespace
Sage Weil [Thu, 8 Apr 2010 15:08:17 +0000 (08:08 -0700)]
osd: fix transposed clone_range args
We're cloning ranges _into_ soid (and in fact just deleted any old soid
that may have been lingering).
Sage Weil [Thu, 8 Apr 2010 14:43:23 +0000 (07:43 -0700)]
spinlock: avoid lockdep include unless lockdep is on
Sage Weil [Thu, 8 Apr 2010 14:43:12 +0000 (07:43 -0700)]
buffer: avoid spinlock include unless debugging is on
Sage Weil [Thu, 8 Apr 2010 14:42:54 +0000 (07:42 -0700)]
librados: reduce #includes
Sage Weil [Wed, 7 Apr 2010 23:43:22 +0000 (16:43 -0700)]
librados: c binding for rados_tmap_update()
Sage Weil [Wed, 7 Apr 2010 23:37:21 +0000 (16:37 -0700)]
rbdtool: maintain an rbd_directory object; add --list
Sage Weil [Wed, 7 Apr 2010 23:35:39 +0000 (16:35 -0700)]
auth: shut up about loading the key file
Sage Weil [Wed, 7 Apr 2010 23:35:28 +0000 (16:35 -0700)]
librados: add tmap_update to c++ interface
Sage Weil [Wed, 7 Apr 2010 23:35:06 +0000 (16:35 -0700)]
osd: fix tmap when header update not specified, or initial object dne
Sage Weil [Wed, 7 Apr 2010 22:29:38 +0000 (15:29 -0700)]
Merge branch 'rbd' into unstable
Sage Weil [Wed, 7 Apr 2010 22:29:23 +0000 (15:29 -0700)]
debian: include rbdtool
Sage Weil [Wed, 7 Apr 2010 22:29:18 +0000 (15:29 -0700)]
rbdtool: add man page
Sage Weil [Wed, 7 Apr 2010 22:29:08 +0000 (15:29 -0700)]
rbdtool: delete rbd images too
Sage Weil [Wed, 7 Apr 2010 21:47:04 +0000 (14:47 -0700)]
rbdtool: use shared rbd_types.h
Sage Weil [Wed, 7 Apr 2010 21:22:25 +0000 (14:22 -0700)]
msgr: drop useless was_session_reset() call on lossy channel close
The fail closes out the pipe, no need to do a 'reset'.
Sage Weil [Wed, 7 Apr 2010 21:16:21 +0000 (14:16 -0700)]
ceph: print output to stdout, not dout
Using cout instead of dout(0) avoids the thread id in the prefix.
Sage Weil [Wed, 7 Apr 2010 21:12:43 +0000 (14:12 -0700)]
msgr: fail on sd < 0
I saw a case where poll(2) was blocking despite being passed an fd of -1.
Since that's clearly invalid, we can fail tcp_{read,write} before that
point.
Weird.
Sage Weil [Wed, 7 Apr 2010 19:33:15 +0000 (12:33 -0700)]
osd: flush store ops after map update
Otherwise we can race:
- queue delete+rewrite of pg log
- unblock op queue
- queue append to log
...
- apply append
- apply delete+rewrite
Sage Weil [Tue, 6 Apr 2010 20:24:58 +0000 (13:24 -0700)]
mds: set laggy in tick() if its been a while
Sage Weil [Tue, 6 Apr 2010 23:18:10 +0000 (16:18 -0700)]
debian: drop empty dir from radosgw
Sage Weil [Tue, 6 Apr 2010 21:54:22 +0000 (14:54 -0700)]
todo: mds
Sage Weil [Tue, 6 Apr 2010 20:25:32 +0000 (13:25 -0700)]
mon: set osd up_thru when adjusting pg_temp
This just avoids making the OSD ask again for up_thru. Fewer osdmap
updates!
Sage Weil [Mon, 15 Mar 2010 17:52:50 +0000 (10:52 -0700)]
osd: distinguish between per-pool snap pools and user-managed snap pools
If removed_snaps is non-empty, then snaps are managed by the
user: snap context is specified for all writes (e.g., MDS or
librados user using the snap context api).
We can enforce this by adding an (unused) snapid (1) to the
removed_snaps the first time a non-pool snap snapid is allocated.
Sage Weil [Tue, 6 Apr 2010 19:29:10 +0000 (12:29 -0700)]
mds: remove null clean stray dentries after dir commit
Sage Weil [Tue, 6 Apr 2010 19:28:43 +0000 (12:28 -0700)]
mds: fix purge_stray to be smart about NEW dentries
Mark new stray dentries NEW.
After purge, remove new dentries immediately; leave others as NULL and
dirty so they get removed from the ondisk stray dir properly.
Sage Weil [Tue, 6 Apr 2010 18:22:29 +0000 (11:22 -0700)]
mds: fix ref count check in eval_stray, purge_stray_purged
Dentry gets pin for any inode pins, and on purge has PIN_PURGING.
Sage Weil [Tue, 6 Apr 2010 18:13:00 +0000 (11:13 -0700)]
osd: return more useful errno (ENOENT) on missing base
Sage Weil [Tue, 6 Apr 2010 17:47:02 +0000 (10:47 -0700)]
mds: move new stray dentries to end of lru
Sage Weil [Tue, 6 Apr 2010 17:46:07 +0000 (10:46 -0700)]
mds: use send_dentry_unlink helper for remote unlink
Sage Weil [Mon, 5 Apr 2010 22:38:45 +0000 (15:38 -0700)]
mds: journal oldest client tid
Journal the client's safe tid with new requests. This keeps the client
completed_requests list trimmed, so that we don't build up a ginormous
list of all requests over the entire journal.
Forward rolling mds format change.
Sage Weil [Fri, 2 Apr 2010 23:41:56 +0000 (16:41 -0700)]
todo: updates
Greg Farnum [Mon, 5 Apr 2010 20:59:52 +0000 (13:59 -0700)]
rgw: rgw_admin checks that it hasn't regenerated user ids
Greg Farnum [Mon, 5 Apr 2010 20:41:37 +0000 (13:41 -0700)]
mon: add new command 'lspools [uid]' to OSDMonitor
Greg Farnum [Mon, 5 Apr 2010 19:50:41 +0000 (12:50 -0700)]
osd: pg_pool_t prints out its owner
Sage Weil [Mon, 5 Apr 2010 17:53:17 +0000 (10:53 -0700)]
mds: fix snap purging during commit
Use the helper!
The commit_partial is still incomplete.. we need to purge the
dentry after the commit completes?
Sage Weil [Mon, 5 Apr 2010 17:26:42 +0000 (10:26 -0700)]
mds: streamline stray checks with helper
Sage Weil [Fri, 2 Apr 2010 22:47:45 +0000 (15:47 -0700)]
mds: make beacon output more informative
Sage Weil [Fri, 2 Apr 2010 22:47:25 +0000 (15:47 -0700)]
qa: run tests in random order
Sage Weil [Fri, 2 Apr 2010 20:34:11 +0000 (13:34 -0700)]
mds: allow reconnect as long as we are in reconnect state
This used to work.. I think this is fallout from the mds session state
simplification (CLOSED was != NEW before).
Anton Blanchard [Sat, 3 Apr 2010 01:51:06 +0000 (12:51 +1100)]
Add missing Makefile dependency for debug.cc
We were missing a Makefile dependency for debug.cc, without it make -j fails
because ceph_ver.h hasn't been created.
Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Sage Weil <sage@newdream.net>
Greg Farnum [Fri, 2 Apr 2010 23:25:00 +0000 (16:25 -0700)]
rgw: radosgw_admin can delete users from the system
Greg Farnum [Fri, 2 Apr 2010 21:21:16 +0000 (14:21 -0700)]
rgw: call url_decode in more appropriate places
Greg Farnum [Fri, 2 Apr 2010 20:14:12 +0000 (13:14 -0700)]
rgw: convert + to space in url_decode
Greg Farnum [Wed, 31 Mar 2010 23:06:45 +0000 (16:06 -0700)]
rgw: send an error response if reading an ACL off disk fails
Greg Farnum [Wed, 31 Mar 2010 18:32:40 +0000 (11:32 -0700)]
rgw: remove bad local variable shadowing class member variable
Greg Farnum [Wed, 31 Mar 2010 18:32:08 +0000 (11:32 -0700)]
rgw: remove some invalid FIXMEs
Yehuda Sadeh [Fri, 2 Apr 2010 21:19:44 +0000 (14:19 -0700)]
rbdtool: don't create empty object data
Yehuda Sadeh [Thu, 1 Apr 2010 00:24:49 +0000 (17:24 -0700)]
rbd: created a userspace tool for creating rbd objects
Sage Weil [Thu, 1 Apr 2010 23:13:41 +0000 (16:13 -0700)]
mds: set laggy flag if dispatch takes too long
This prevents, e.g., tick() from coming in immediately after and
closing out client sessions after we chew through something particularly
slow.
Sage Weil [Thu, 1 Apr 2010 20:46:26 +0000 (13:46 -0700)]
msgr: set OPEN state after accepting connection
Not doing so can eventually lead to
msg/SimpleMessenger.cc: In function 'int SimpleMessenger::Pipe::accept()':
msg/SimpleMessenger.cc:765: FAILED assert(existing->state == STATE_CONNECTING)
Sage Weil [Thu, 1 Apr 2010 20:06:51 +0000 (13:06 -0700)]
mds: fix file recovery
This was broken by commit
d5574993f .
Sage Weil [Thu, 1 Apr 2010 20:03:04 +0000 (13:03 -0700)]
mds: remove dir from 'new' list on any commit, not just on clean
A dir may be redirtied after the commit, such that it never becomes clean.
It only needs to stay on the 'new' list until it's been written to disk
at least once, though.
Sage Weil [Thu, 1 Apr 2010 18:39:16 +0000 (11:39 -0700)]
filejournal: fix journal full handling
We would block on journal full, but then try to continue where we left off,
which broke thoroughly. Add return codes, and wait at the proper times:
- if journal if full on first event, wait. otherwise, write what we have
so far.
- wait in write_thread_entry(), not check_for_full().
Also fix up 'room' calculation.
Sage Weil [Thu, 1 Apr 2010 16:44:04 +0000 (09:44 -0700)]
qa: make snaptest trigger capsnap dirty(caps)==0 bug
Prior to ceph-client.git commit
819ccbfa448 , this will cause a
VFS: Busy inodes after unmount of ceph. Self-destruct in 5 seconds. Have a nice day...
on umount.
Sage Weil [Thu, 1 Apr 2010 14:34:24 +0000 (07:34 -0700)]
mds: remove dentry AND inode when dropping snap metadata; add helper
We should only drop obsolete snapped metadata when it is unreferenced, and
at that point we need to drop the dentry AND inode, not just the dentry.
This delays things until caps are released, among other things.
Sage Weil [Thu, 1 Apr 2010 14:28:41 +0000 (07:28 -0700)]
osd: improve print of snapc in MOSDOp
Sage Weil [Thu, 1 Apr 2010 14:27:59 +0000 (07:27 -0700)]
osd: fix snap trimmer SnapSet adjustment
AFAICS it's totally wrong?
Sage Weil [Wed, 31 Mar 2010 22:30:29 +0000 (15:30 -0700)]
todo: updates
Sage Weil [Wed, 31 Mar 2010 22:30:17 +0000 (15:30 -0700)]
osd: improve snap trimmer debug output
Sage Weil [Wed, 31 Mar 2010 22:00:23 +0000 (15:00 -0700)]
mds: clear old snap_trace before building new one
Otherwise we just build up a big stack of historical information and send
it all to the client.
Sage Weil [Wed, 31 Mar 2010 21:46:20 +0000 (14:46 -0700)]
qa: make snaptest.sh repeatable in same dir
Sage Weil [Wed, 31 Mar 2010 16:49:55 +0000 (09:49 -0700)]
Merge branch 'mds' into unstable
Sage Weil [Wed, 31 Mar 2010 04:27:50 +0000 (21:27 -0700)]
todo: kclient snap bug
Sage Weil [Wed, 31 Mar 2010 04:27:14 +0000 (21:27 -0700)]
mds: don't adjust subtree map in rename_prepare
Not sure what the reasoning behind this was.
This code is from pre git history, and the git->subversion conversion
managed to make pretty git-blame unusable. I doubt I really documented
what its purpose at that point was anyway.
Sage Weil [Wed, 31 Mar 2010 04:25:46 +0000 (21:25 -0700)]
mds: subtree map should always reflect current hierarchy
Not projected!
Sage Weil [Wed, 31 Mar 2010 04:24:45 +0000 (21:24 -0700)]
qa: drop numeric prefix from workunits
Sage Weil [Tue, 30 Mar 2010 20:40:41 +0000 (13:40 -0700)]
osd: don't filter out shutdown msg to self
Sage Weil [Tue, 30 Mar 2010 20:19:30 +0000 (13:19 -0700)]
msgr: init local_pipe peer_{addr,type}
Sage Weil [Tue, 30 Mar 2010 19:22:37 +0000 (12:22 -0700)]
Merge branch 'mds' into unstable
Sage Weil [Tue, 30 Mar 2010 19:14:03 +0000 (12:14 -0700)]
mds: todo rejoin bug
Sage Weil [Tue, 30 Mar 2010 19:13:33 +0000 (12:13 -0700)]
todo: kclient bug
Greg Farnum [Tue, 30 Mar 2010 18:48:06 +0000 (11:48 -0700)]
rgw: revert accidental commit of get_attr prototype change
Sage Weil [Tue, 30 Mar 2010 18:45:00 +0000 (11:45 -0700)]
Merge branch 'msgr' into unstable
Sage Weil [Fri, 26 Mar 2010 19:31:39 +0000 (12:31 -0700)]
msgr: source, not orig_source, in dbg output
Sage Weil [Thu, 25 Mar 2010 23:09:42 +0000 (16:09 -0700)]
mds: don't use get_orig_source
MDS doesn't do any request forwarding, so there is no need.
Sage Weil [Thu, 25 Mar 2010 23:04:18 +0000 (16:04 -0700)]
msgr: NOSRCADDR protocol feature drops src/orig_src in msg header
This lets us drop the src, orig_src ceph_entity_addr's from the message
header, saving about 160 bytes per message.
The feature is optional. We can still talk to peers who use the old
protocol.
Sage Weil [Thu, 25 Mar 2010 23:01:34 +0000 (16:01 -0700)]
msgr: put features in connection_state
Sage Weil [Thu, 25 Mar 2010 21:49:56 +0000 (14:49 -0700)]
msgr: use connection for src addr
Sage Weil [Thu, 25 Mar 2010 20:46:16 +0000 (13:46 -0700)]
msgr: make Message::get_orig_* differ only when explicitly directed to
This paves the way for removal of the orig_src field from the message
header.
Sage Weil [Tue, 30 Mar 2010 17:30:02 +0000 (10:30 -0700)]
mds: fix MDSTableClient ack double journaling
Do not journal ack unless the tid is registered in the LogSegment. Once
we journal it, we remove it from the LogSegment list, and once it's
journaled, we remove the pending_commit[tid] entry.
This fixes a bug where the mds got two acks, journaled both of them, and
crashed in the completion for the second because pending_commit[tid] was
gone. The second ack should have been ignored.
Sage Weil [Mon, 29 Mar 2010 23:26:54 +0000 (16:26 -0700)]
mds: start file recovery after sending rejoin ack
The rejoin ack intializes replica lock states correctly; we can't send any
lock messages before that. This fixes both the check max size call (which
sends lock messages taking the wrlock) and the file_recover() call
(which does the same).
Instead, we make two lists, files to recover and those to fix up. The lock
states for both are set to PRE_SCAN (LOCK on replica). After the rejoin
acks go out, we either check_inode_max_size or file_recover.
If file_recover someday grows another caller, this may need something a bit
more sophisticated.
Sage Weil [Mon, 29 Mar 2010 22:52:51 +0000 (15:52 -0700)]
mds: delay check_inode_max_size if frozen
Sage Weil [Mon, 29 Mar 2010 21:34:52 +0000 (14:34 -0700)]
mds: less noisy about snaprealm siblings
This can get _very_ large for the stray dir.
Greg Farnum [Mon, 29 Mar 2010 18:43:45 +0000 (11:43 -0700)]
rgw: add optional auid parameter to create_bucket.
If set this will set the RADOS pool's auid to the given one, granting
partial data coherency.
Greg Farnum [Mon, 29 Mar 2010 17:03:29 +0000 (10:03 -0700)]
rgw: remove dead code. rebuild_policy exists elsewhere; the acl needs no main()
Greg Farnum [Mon, 29 Mar 2010 14:40:51 +0000 (07:40 -0700)]
rgw: radosgw_admin will quit gracefully if it can't open the store
Greg Farnum [Mon, 29 Mar 2010 14:40:28 +0000 (07:40 -0700)]
rgw: set auid if specified at creation
Greg Farnum [Fri, 26 Mar 2010 23:29:11 +0000 (16:29 -0700)]
rgw: add auid field to RGWUserInfo
Greg Farnum [Fri, 26 Mar 2010 22:57:14 +0000 (15:57 -0700)]
auth: Keyring no longer segfaults if HOME env variable is unset.
Just skip that path, instead. We can't recover what it meant.
Greg Farnum [Fri, 26 Mar 2010 22:56:37 +0000 (15:56 -0700)]
rgw: if you can't init_storage_provider(), quit instead
of segfaulting later.
Greg Farnum [Fri, 26 Mar 2010 22:56:14 +0000 (15:56 -0700)]
rgw: say why the store failed to open in admin
Greg Farnum [Fri, 26 Mar 2010 22:55:47 +0000 (15:55 -0700)]
rgw: add minor data to schema examination
Sage Weil [Mon, 29 Mar 2010 16:48:28 +0000 (09:48 -0700)]
mkcephfs: osd=allow * caps for osd, mds
Sage Weil [Fri, 26 Mar 2010 23:01:35 +0000 (16:01 -0700)]
mds: migrate frag/nest scatterlock info on bounding frags during export
This ensures that the auth inode continues to maintain accurate scatterlock
info about open frags. We include info on export if it is a bounding frag.
On import, we only take it if we are !auth. This mirrors the scatterlock
scatter/gather logic in CInode::{encode,decode}_lock_state.
Sage Weil [Fri, 26 Mar 2010 23:04:26 +0000 (16:04 -0700)]
auth: standardize on -n/--name [type.]name
Sage Weil [Fri, 26 Mar 2010 22:46:52 +0000 (15:46 -0700)]
auth: initialized uninitialized variable, don't free pts we don't own
Sage Weil [Fri, 26 Mar 2010 21:12:42 +0000 (14:12 -0700)]
mds: make rdlock waiter wait on STABLE | RD, not just RD, when lock is unstable
Sage Weil [Fri, 26 Mar 2010 20:35:42 +0000 (13:35 -0700)]
mds: use 32 bits for MDSCacheObject::ref count
16 is not enough! :)
Sage Weil [Fri, 26 Mar 2010 20:22:04 +0000 (13:22 -0700)]
Merge branch 'unstable' into mds
Conflicts:
src/TODO