]>
git.apps.os.sepia.ceph.com Git - ceph.git/log
Sage Weil [Thu, 24 Sep 2009 21:56:39 +0000 (14:56 -0700)]
objecter: optionally set priority on ObjectOperations
This just sets the priority on the MOSDOp we send.
Sage Weil [Thu, 24 Sep 2009 21:10:57 +0000 (14:10 -0700)]
mds: do not attach mdr to NULL session
Sage Weil [Wed, 23 Sep 2009 17:11:56 +0000 (10:11 -0700)]
todo
Sage Weil [Wed, 23 Sep 2009 17:00:32 +0000 (10:00 -0700)]
mds: warn on dup inodes; don't crash
We just ignore the second instance of the inode. It can get
cleaned up transparently if/when the directory gets saved.
Sage Weil [Tue, 22 Sep 2009 22:12:18 +0000 (15:12 -0700)]
objecter: remove dead code
Sage Weil [Tue, 22 Sep 2009 22:11:47 +0000 (15:11 -0700)]
librados: pass resets to objecter, with proper locking
Sage Weil [Tue, 22 Sep 2009 22:11:34 +0000 (15:11 -0700)]
mds: pass resets to objecter, with proper locking
Sage Weil [Tue, 22 Sep 2009 22:11:25 +0000 (15:11 -0700)]
uclient: pass reset events to objecter, with proper locking
Sage Weil [Tue, 22 Sep 2009 22:11:10 +0000 (15:11 -0700)]
objecter: resend mon ops on handle mon connection reset
Sage Weil [Tue, 22 Sep 2009 22:37:51 +0000 (15:37 -0700)]
mds: fix inode load existence check
We can lookup using last snapid, not first. That's how the
inode_map is indexed.
Sage Weil [Tue, 22 Sep 2009 21:18:47 +0000 (14:18 -0700)]
todo
Sage Weil [Tue, 22 Sep 2009 21:17:50 +0000 (14:17 -0700)]
uclient: don't readdir first frag on opendir
This reorders things such that rm -r breaks. It was an old old
optimization that doesn't make sense anymore anyway.
Sage Weil [Tue, 22 Sep 2009 20:17:03 +0000 (13:17 -0700)]
objecter: fix uninitialized variable last_seen_version
Fixes hang on 'rados df'
Sage Weil [Tue, 22 Sep 2009 20:09:09 +0000 (13:09 -0700)]
librados: fix dispatch
Greg Farnum [Tue, 22 Sep 2009 00:27:59 +0000 (17:27 -0700)]
uclient: Better fix for insert_dentry_inode to work with pre-built Dentrys.
Greg Farnum [Tue, 22 Sep 2009 00:07:20 +0000 (17:07 -0700)]
uclient: get_or_create should return out-of-date lease Dentrys too
Greg Farnum [Tue, 22 Sep 2009 00:06:14 +0000 (17:06 -0700)]
uclient: encode_cap_releases shouldn't update caps if inode isn't in the cache
Greg Farnum [Tue, 22 Sep 2009 00:04:32 +0000 (17:04 -0700)]
uclient: First attempt at preemptive cache dropping in the functions.
Greg Farnum [Mon, 21 Sep 2009 21:13:19 +0000 (14:13 -0700)]
uclient: Don't *always* set_dentry_wanted() on MClientRequests.
Greg Farnum [Mon, 21 Sep 2009 20:46:37 +0000 (13:46 -0700)]
uclient: Only make the MClientRequest in send_request itself.
Add a few more functions to MetaRequest so we can use those in choose_mds.
Greg Farnum [Mon, 21 Sep 2009 20:11:52 +0000 (13:11 -0700)]
uclient: More debugging. May want to back off the inode printout to
not include caps_issued.
Greg Farnum [Mon, 21 Sep 2009 18:13:55 +0000 (11:13 -0700)]
uclient: changed link and insert_dentry_inode to behave with pre-built dentrys
Greg Farnum [Fri, 18 Sep 2009 22:44:45 +0000 (15:44 -0700)]
uclient: Attempt preemptive cap dropping on mknod.
Greg Farnum [Fri, 18 Sep 2009 21:21:58 +0000 (14:21 -0700)]
uclient: dout statements in new functions.
Greg Farnum [Fri, 18 Sep 2009 22:20:29 +0000 (15:20 -0700)]
uclient: Fixes for get_or_create
Greg Farnum [Fri, 18 Sep 2009 20:51:26 +0000 (13:51 -0700)]
uclient: Functions to create Dentrys prior to gettin caps from mds.
Sage Weil [Tue, 22 Sep 2009 19:29:33 +0000 (12:29 -0700)]
remove extra whitespace
Sage Weil [Tue, 22 Sep 2009 19:26:52 +0000 (12:26 -0700)]
kclient: fix bit test logic
This fixes the fault messages with <NULL> in them; we now print
the proper message, and in a timely fashion.
Sage Weil [Tue, 22 Sep 2009 19:15:33 +0000 (12:15 -0700)]
kclient: use pr_fmt to prefix module name
This doubles up the ceph:ceph: if we have CONFIG_DYNAMIC_DEBUG, but
that's not our fault.
Sage Weil [Tue, 22 Sep 2009 18:50:29 +0000 (11:50 -0700)]
kclient: prefix frag_ functions with ceph_
It was suggested this would avoid any potential conflict with TCP.
It's a good idea to keep the namespace clean anyway.
Sage Weil [Tue, 22 Sep 2009 18:27:32 +0000 (11:27 -0700)]
kclient: kill out_qlen
This is apparently buggy (not immediately obvious why) but it's
only really useful for debugging anyway (it makes it easier to
match up message sequence numbers). Rip it out.
Sage Weil [Tue, 22 Sep 2009 18:17:22 +0000 (11:17 -0700)]
make inline string functions non-inline
Sage Weil [Tue, 22 Sep 2009 17:17:40 +0000 (10:17 -0700)]
kclient: whitespace
Sage Weil [Tue, 22 Sep 2009 17:13:11 +0000 (10:13 -0700)]
v0.15
Sage Weil [Tue, 22 Sep 2009 17:10:40 +0000 (10:10 -0700)]
Makefile: include ioctl-number.patch
Sage Weil [Tue, 22 Sep 2009 17:10:27 +0000 (10:10 -0700)]
msgr: don't print source in msg debug output
Sage Weil [Tue, 22 Sep 2009 17:00:31 +0000 (10:00 -0700)]
todo
Sage Weil [Tue, 22 Sep 2009 16:59:23 +0000 (09:59 -0700)]
monclient: change pick_new_mon -> reopen_session
Sage Weil [Tue, 22 Sep 2009 16:58:07 +0000 (09:58 -0700)]
monclient: avoid dup subscribe messages
Sage Weil [Tue, 22 Sep 2009 16:55:21 +0000 (09:55 -0700)]
osd: subscribe to monmap
Sage Weil [Mon, 21 Sep 2009 23:52:40 +0000 (16:52 -0700)]
todo
Sage Weil [Mon, 21 Sep 2009 23:52:19 +0000 (16:52 -0700)]
kclient: fix le encoding bugs (yay sparse)
Sage Weil [Mon, 21 Sep 2009 23:49:56 +0000 (16:49 -0700)]
kclient: clean whitespace
Sage Weil [Mon, 21 Sep 2009 23:40:00 +0000 (16:40 -0700)]
kclient: include object name in DATALOC ioctl result
Sage Weil [Mon, 21 Sep 2009 23:26:50 +0000 (16:26 -0700)]
kclient: include ioctl-number doc file in patch set
Sage Weil [Mon, 21 Sep 2009 23:26:35 +0000 (16:26 -0700)]
kclient: no need for static ceph_buffer
Sage Weil [Mon, 21 Sep 2009 22:11:07 +0000 (15:11 -0700)]
mds: fix mds standby map handling
Borken by
c6ddbdb4671b3594f9b4d5f54b65f8525935758b
Sage Weil [Mon, 21 Sep 2009 21:55:59 +0000 (14:55 -0700)]
mon: consolidate sub request msgs; cleaner api
Caller must call renew_subs() explicitly when they want to start
the subscription(s).
Sage Weil [Mon, 21 Sep 2009 21:45:45 +0000 (14:45 -0700)]
mon: allow sub for monmap
Sage Weil [Mon, 21 Sep 2009 21:30:19 +0000 (14:30 -0700)]
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.
Sage Weil [Mon, 21 Sep 2009 21:24:35 +0000 (14:24 -0700)]
msgr: pass Connection* to msgr error handlers
This lets the error handlers get at session state (if any).
Sage Weil [Mon, 21 Sep 2009 21:06:15 +0000 (14:06 -0700)]
mon: fix client id preallocation
Broken by
b2c7655093821710aac2b4d35ccb1740429eb402
Sage Weil [Mon, 21 Sep 2009 21:05:57 +0000 (14:05 -0700)]
monclient: send a single mount request
Sage Weil [Fri, 18 Sep 2009 04:41:04 +0000 (21:41 -0700)]
uclient: use ref in constructor for MClientRequest
Sage Weil [Fri, 18 Sep 2009 23:41:38 +0000 (16:41 -0700)]
kclient: two msgpool modes: blocking (ala mempool_t) and !blocking (strict)
Only warn on !blocking pool depletion!
Sage Weil [Fri, 18 Sep 2009 23:32:26 +0000 (16:32 -0700)]
kclient: fix symlink kmalloc under spinlock
Sage Weil [Fri, 18 Sep 2009 23:17:04 +0000 (16:17 -0700)]
strerror -> strerror_r
Sage Weil [Fri, 18 Sep 2009 21:57:35 +0000 (14:57 -0700)]
kclient: rename msgpool type
Sage Weil [Fri, 18 Sep 2009 21:54:04 +0000 (14:54 -0700)]
kclient: use msgpools for osd requests, replies
Sage Weil [Fri, 18 Sep 2009 21:18:43 +0000 (14:18 -0700)]
uclient: fix null pointer ref
Sage Weil [Fri, 18 Sep 2009 21:15:44 +0000 (14:15 -0700)]
kclient: fix up ceph_client init error handling
Sage Weil [Fri, 18 Sep 2009 21:05:10 +0000 (14:05 -0700)]
kclient: fix patchset comment
Sage Weil [Fri, 18 Sep 2009 21:04:50 +0000 (14:04 -0700)]
kclient: fall back to WARN + new allocation if msgpool is depleted
Sage Weil [Fri, 18 Sep 2009 20:08:25 +0000 (13:08 -0700)]
kclient: use msgpools for other monc messages
Sage Weil [Fri, 18 Sep 2009 20:01:22 +0000 (13:01 -0700)]
kclient: add back msgpools
This reverts commit
b985882aa27fea1d1ed92b09b6057be961c56c0a .
Sage Weil [Fri, 18 Sep 2009 19:45:48 +0000 (12:45 -0700)]
kclient: add GET_DATALOC ioctl to get object location, osd addr
Greg Farnum [Fri, 18 Sep 2009 00:25:04 +0000 (17:25 -0700)]
uclient: _removexattr should be naming the xattr to remove!
Greg Farnum [Fri, 18 Sep 2009 00:22:42 +0000 (17:22 -0700)]
uclient: Add functions/data members for preemptive cap dropping.
Greg Farnum [Thu, 17 Sep 2009 18:16:33 +0000 (11:16 -0700)]
MClientRequest: Releases should have a constructor
Greg Farnum [Wed, 16 Sep 2009 00:32:01 +0000 (17:32 -0700)]
uclient: First pass at encode_cap_release function; not used.
Greg Farnum [Tue, 15 Sep 2009 20:08:25 +0000 (13:08 -0700)]
uclient: Move the MetaRequest to MClientRequest into its own code, and
cut out the payload encode/decode since MClientRequests are now made
from MetaRequests.
Oh, and there are no issues with rebuilding messages -- false alarm!
Greg Farnum [Tue, 15 Sep 2009 18:40:52 +0000 (11:40 -0700)]
uclient: The basic passing mechanism is now a MetaRequest. It seems to work.
More bug-testing is needed, though, as rebuilding messages in send_message
is finicky at best...
Greg Farnum [Mon, 14 Sep 2009 23:31:35 +0000 (16:31 -0700)]
uclient: Make the uclient work properly with the new MetaRequest structure.
Greg Farnum [Mon, 14 Sep 2009 22:40:14 +0000 (15:40 -0700)]
uclient: MetaRequests are going to be how we pass requests around;
repurpose them so they don't require an MClientRequest.
ALSO, convert the mds member from a map to an int.
Greg Farnum [Mon, 14 Sep 2009 17:46:39 +0000 (10:46 -0700)]
Hadoop: Minor fixes and style changes to satisfy Hadoop's FindBugs system.
Remove a few deprecated methods.
Sage Weil [Thu, 17 Sep 2009 23:15:21 +0000 (16:15 -0700)]
kclient: fix osd request kicking
We often need to kick the request to the same osd. Keep track of
an osd incarnation, incremented each time we reset_osd or see the
connection drop, and kick requests when the incarnation doesn't
match.
Sage Weil [Thu, 17 Sep 2009 22:57:06 +0000 (15:57 -0700)]
mon: fix bug with client id assignment of -1
Sage Weil [Thu, 17 Sep 2009 22:56:43 +0000 (15:56 -0700)]
kclient: avoid no-op dir cap release when no dentry lease
Make release logic a bit smarter: only force inclusion of dir
cap release if there is a dentry lease to start with.
Sage Weil [Thu, 17 Sep 2009 22:29:09 +0000 (15:29 -0700)]
kclient: make mds console msgs consistent
Sage Weil [Thu, 17 Sep 2009 21:56:34 +0000 (14:56 -0700)]
monc: pick monitor before sending initial mount
Sage Weil [Thu, 17 Sep 2009 21:54:36 +0000 (14:54 -0700)]
uclient: implement basic I_COMPLETE support
Sage Weil [Thu, 17 Sep 2009 21:25:55 +0000 (14:25 -0700)]
uclient: factor out command cap issue checks
rdcache_gen, shared_gen
Sage Weil [Thu, 17 Sep 2009 21:16:05 +0000 (14:16 -0700)]
mds: minor comments
Sage Weil [Thu, 17 Sep 2009 21:15:53 +0000 (14:15 -0700)]
kclient: move some inode update code inside i_lock
Sage Weil [Thu, 17 Sep 2009 21:15:39 +0000 (14:15 -0700)]
kclient: use shared_gen, not cache_gen, for I_COMPLETE tracking
Sage Weil [Thu, 17 Sep 2009 20:47:35 +0000 (13:47 -0700)]
todo
Sage Weil [Wed, 16 Sep 2009 19:12:43 +0000 (12:12 -0700)]
monclient: small cleanup
Sage Weil [Wed, 16 Sep 2009 19:12:34 +0000 (12:12 -0700)]
todo
Sage Weil [Thu, 17 Sep 2009 20:15:29 +0000 (13:15 -0700)]
kclient: fix console msg
Sage Weil [Thu, 17 Sep 2009 19:04:28 +0000 (12:04 -0700)]
kclient: move caps stats into client debugfs dir; fix error checking
Sage Weil [Thu, 17 Sep 2009 18:28:53 +0000 (11:28 -0700)]
kclient: put fsid in debugfs dir name
Sage Weil [Thu, 17 Sep 2009 18:58:06 +0000 (11:58 -0700)]
mds: make parallel_fetch work with hard/remote links
We need to open remote ino's, and possibly add to gather (like
a fetch). Required some refactoring. Killed/renamed
otherwise unused traverse_to_dir.
Sage Weil [Thu, 17 Sep 2009 17:50:14 +0000 (10:50 -0700)]
cosd.conf: fix date stamp
Sage Weil [Thu, 17 Sep 2009 17:41:07 +0000 (10:41 -0700)]
msgr: fix null deref setting message connection ref
Sage Weil [Wed, 16 Sep 2009 17:34:04 +0000 (10:34 -0700)]
configure: don't link expat, fcgi to everything
The default AC_CHECK_LIB action is to add -lfoo to the linker args,
which we don't want!
Sage Weil [Wed, 16 Sep 2009 04:27:41 +0000 (21:27 -0700)]
Makefile: fix mode on installed headers
Sage Weil [Tue, 15 Sep 2009 22:02:17 +0000 (15:02 -0700)]
msgr: change dispatcher interfaces; some monclient cleanups
Sage Weil [Tue, 15 Sep 2009 20:14:01 +0000 (13:14 -0700)]
uclient: do not do locking for 'atomic' writes
They aren't tested and aren't safe (client may fail; there's no cleanup).
Sage Weil [Mon, 14 Sep 2009 22:34:47 +0000 (15:34 -0700)]
mds: flush log in caller(s), not scatter_writebehind
Otherwise scatter_tick(), in particular, will call scatter_writebehind
a zillion times and generate lots of small log flushes. Instead, flush
at the end of the whole tick.
Sage Weil [Mon, 14 Sep 2009 22:07:30 +0000 (15:07 -0700)]
check_mds_request_latency: fix parsing for current debug output