]> git.apps.os.sepia.ceph.com Git - ceph.git/log
ceph.git
15 years agoTODO
Greg Farnum [Fri, 8 Jan 2010 23:42:36 +0000 (15:42 -0800)]
TODO

15 years agomon: properly clean up SessionMap on a shutdown.
Greg Farnum [Sat, 9 Jan 2010 01:25:53 +0000 (17:25 -0800)]
mon: properly clean up SessionMap on a shutdown.

This fixes failed asserts whenever you tried to do a clean monitor shutdown
instead of killing it.

15 years agocsyn: properly use multiple clients
Greg Farnum [Fri, 8 Jan 2010 21:42:46 +0000 (13:42 -0800)]
csyn: properly use multiple clients

15 years agodon't keep configs in git
Sage Weil [Fri, 8 Jan 2010 19:19:23 +0000 (11:19 -0800)]
don't keep configs in git

15 years agotodo
Sage Weil [Fri, 8 Jan 2010 18:13:50 +0000 (10:13 -0800)]
todo

15 years agobotched merge
Greg Farnum [Fri, 8 Jan 2010 01:35:55 +0000 (17:35 -0800)]
botched merge

15 years agoMerge branch 'unstable' of ceph.newdream.net:/git/ceph into unstable
Greg Farnum [Fri, 8 Jan 2010 01:15:39 +0000 (17:15 -0800)]
Merge branch 'unstable' of ceph.newdream.net:/git/ceph into unstable

Conflicts:

src/msg/SimpleMessenger.cc

15 years agomsgr: make use of references and const slightly more consistent
Greg Farnum [Fri, 8 Jan 2010 01:10:37 +0000 (17:10 -0800)]
msgr: make use of references and const slightly more consistent

15 years agomonClient: fix potential naming confusion with messenger
Greg Farnum [Fri, 8 Jan 2010 01:10:21 +0000 (17:10 -0800)]
monClient: fix potential naming confusion with messenger

15 years agomsgr: rename endpoint_stopped to destination_stopped
Greg Farnum [Fri, 8 Jan 2010 00:56:28 +0000 (16:56 -0800)]
msgr: rename endpoint_stopped to destination_stopped

15 years agomsgr: rank is just a bad name for local SimpleMessengers now
Greg Farnum [Fri, 8 Jan 2010 00:53:51 +0000 (16:53 -0800)]
msgr: rank is just a bad name for local SimpleMessengers now

15 years agomsgr: remove erank
Sage Weil [Fri, 8 Jan 2010 00:10:23 +0000 (16:10 -0800)]
msgr: remove erank

Obsolete, now that we have only a single entity per SimpleMessenger
instance.  Avoid changing protocol by making it a reserved (and ignored)
field.

15 years agomsgr: should be destroyed in the same thread it's created
Greg Farnum [Thu, 7 Jan 2010 23:35:26 +0000 (15:35 -0800)]
msgr: should be destroyed in the same thread it's created

15 years agofilestore: don't bother masking signals
Sage Weil [Thu, 17 Dec 2009 20:11:16 +0000 (12:11 -0800)]
filestore: don't bother masking signals

The OSD already catches SIGINT, SIGTERM.  It's not really
effective anyway.

15 years agomsgr: Remove erroneous rank->destroy calls I inserted in 9a96bf
Greg Farnum [Thu, 7 Jan 2010 20:05:38 +0000 (12:05 -0800)]
msgr: Remove erroneous rank->destroy calls I inserted in 9a96bf

15 years agomsgr: get and put inside DispatchThread, not when you create/end DispatchThread
Greg Farnum [Thu, 7 Jan 2010 01:15:44 +0000 (17:15 -0800)]
msgr: get and put inside DispatchThread, not when you create/end DispatchThread

15 years agomsgr: Remove unregister_entity since there's only ever one now.
Greg Farnum [Thu, 7 Jan 2010 00:14:45 +0000 (16:14 -0800)]
msgr: Remove unregister_entity since there's only ever one now.
Move needed functionality into the end of dispatch_entry.

15 years agomsgr: Make usage compatible with last commit
Greg Farnum [Wed, 6 Jan 2010 22:13:05 +0000 (14:13 -0800)]
msgr: Make usage compatible with last commit

15 years agomsgr: Remove all references to Endpoint
Greg Farnum [Wed, 6 Jan 2010 01:09:41 +0000 (17:09 -0800)]
msgr: Remove all references to Endpoint

15 years agomsgr: Remove declarations of unimplemented functions:
Greg Farnum [Wed, 6 Jan 2010 00:51:01 +0000 (16:51 -0800)]
msgr: Remove declarations of unimplemented functions:
rename_entity and new_entity

15 years agomsgr: SimpleMessenger takes responsibility for Messenger functions.
Greg Farnum [Wed, 6 Jan 2010 00:48:17 +0000 (16:48 -0800)]
msgr: SimpleMessenger takes responsibility for Messenger functions.

15 years agomsgr: All SimpleMessenger uses are now on the heap for appropriate destruction
Greg Farnum [Tue, 5 Jan 2010 20:02:59 +0000 (12:02 -0800)]
msgr: All SimpleMessenger uses are now on the heap for appropriate destruction

15 years agomsgr: SimpleMessenger is now a Messenger itself -- sort of
Greg Farnum [Tue, 5 Jan 2010 00:04:28 +0000 (16:04 -0800)]
msgr: SimpleMessenger is now a Messenger itself -- sort of

15 years agomsgr: remove rank number from endpoint, since there's only one.
Greg Farnum [Mon, 4 Jan 2010 18:01:41 +0000 (10:01 -0800)]
msgr: remove rank number from endpoint, since there's only one.
Remove Endpoint's wait() method as it's unused.

15 years agotodo
Sage Weil [Thu, 7 Jan 2010 00:08:29 +0000 (16:08 -0800)]
todo

15 years agoosd: fix MASKTRUNC
Sage Weil [Thu, 7 Jan 2010 00:03:31 +0000 (16:03 -0800)]
osd: fix MASKTRUNC

This was looking at the wrong osd_op for the truncate_seq, which made it
essentially broken.  This was problematic any time the client SETTRUNC
reached an object (on a write after truncate) before the MDS TRIMTRUNC did.

15 years agoqa: dl pjd.tgz
Sage Weil [Wed, 23 Dec 2009 22:49:32 +0000 (14:49 -0800)]
qa: dl pjd.tgz

Then we can switch the symlink later

15 years agotodo, sepia reformat
Sage Weil [Wed, 23 Dec 2009 22:49:16 +0000 (14:49 -0800)]
todo, sepia reformat

15 years agomds: change MClientReconnect encoding (mds protocol change)
Sage Weil [Wed, 23 Dec 2009 20:09:41 +0000 (12:09 -0800)]
mds: change MClientReconnect encoding (mds protocol change)

This simplifies the encoding to make life easier for the kclient.

15 years agouclient: clean up mark_down of failed mds
Sage Weil [Wed, 23 Dec 2009 20:04:39 +0000 (12:04 -0800)]
uclient: clean up mark_down of failed mds

Set inst at the time we see it change, otherwise we may miss an mds that
replayed but failed before it fully recovered.

15 years agomds: add NEW session state
Sage Weil [Wed, 23 Dec 2009 19:52:19 +0000 (11:52 -0800)]
mds: add NEW session state

We want to register new sessions in the map so that they can be found when
we replay ESession events in the log.  Otherwise a client that connects
to the mds early will have one Session* on the connection, and another in
the map, and reconnect will fail.

Make sure we remove NEW sessions from the map on connection reset, so that
things gets freed.

Also, take mds_lock in ms_verify_authorizer.

15 years agomsgr: drop rank lock while verifying authorizer
Sage Weil [Wed, 23 Dec 2009 19:49:51 +0000 (11:49 -0800)]
msgr: drop rank lock while verifying authorizer

This easily causes a lock inversion if the verify callback takes any
locks.  e.g.

09.12.23 11:43:32.335042 1482578256 lockdep: ------------------------------------
09.12.23 11:43:32.335072 1482578256 lockdep: existing dependency MDS::mds_lock (10) -> SimpleMessenger::lock (6) at:
 1: (Mutex::Lock(bool)+0x3a) [0x6f581c]
 2: (SimpleMessenger::submit_message(Message*, entity_inst_t const&, bool)+0xc0) [0x6f900c]
 3: (SimpleMessenger::Endpoint::send_message(Message*, entity_inst_t)+0x1ab) [0x6f9c85]
 4: (MonClient::_send_mon_message(Message*, bool)+0xa2) [0x9682dc]
 5: (MonClient::send_mon_message(Message*)+0x37) [0x740e17]
 6: (MDS::beacon_send()+0x226) [0x710b02]
 7: (MDS::beacon_start()+0x15) [0x711fc5]
 8: (MDS::init()+0xd5) [0x71238b]
 9: (main()+0x2cd) [0x6eb6af]
 10: (__libc_start_main()+0xe6) [0x7f22954811a6]
 11: /tmp/cmds.20091223.114259 [0x6eb0d9]

09.12.23 11:43:32.335289 1482578256 lockdep: new dependency SimpleMessenger::lock (6) -> MDS::mds_lock (10) creates a cycle at
 1: (Mutex::Lock(bool)+0x3a) [0x6f581c]
 2: (Mutex::Locker::Locker(Mutex&)+0x2c) [0x705f2c]
 3: (MDS::ms_verify_authorizer(Connection*, int, int, buffer::list&, buffer::list&, bool&)+0x48) [0x70d6c6]
 4: (Messenger::ms_deliver_verify_authorizer(Connection*, int, int, buffer::list&, buffer::list&, bool&)+0x74) [0x702d5e]
 5: (SimpleMessenger::verify_authorizer(Connection*, int, int, buffer::list&, buffer::list&, bool&)+0x6b) [0x6f6279]
 6: (SimpleMessenger::Pipe::accept()+0xc7d) [0x6ffba5]
 7: (SimpleMessenger::Pipe::reader()+0x32) [0x700de8]
 8: (SimpleMessenger::Pipe::Reader::entry()+0x19) [0x6f4e2d]
 9: (Thread::_entry_func(void*)+0x20) [0x705944]
 10: /lib/libpthread.so.0 [0x7f22962effc7]
 11: (clone()+0x6d) [0x7f22955325ad]

09.12.23 11:43:32.335497 1482578256 lockdep: btw, i am holding these locks:
09.12.23 11:43:32.335511 1482578256 lockdep:   SimpleMessenger::lock (6)

common/lockdep.cc: In function 'int lockdep_will_lock(const char*, int)':
common/lockdep.cc:184: FAILED assert(0)
 1: (lockdep_will_lock(char const*, int)+0x765) [0x9776a9]
 2: (Mutex::_will_lock()+0x1f) [0x6f43bf]
 3: (Mutex::Lock(bool)+0x3a) [0x6f581c]
 4: (Mutex::Locker::Locker(Mutex&)+0x2c) [0x705f2c]
 5: (MDS::ms_verify_authorizer(Connection*, int, int, buffer::list&, buffer::list&, bool&)+0x48) [0x70d6c6]
 6: (Messenger::ms_deliver_verify_authorizer(Connection*, int, int, buffer::list&, buffer::list&, bool&)+0x74) [0x702d5e]
 7: (SimpleMessenger::verify_authorizer(Connection*, int, int, buffer::list&, buffer::list&, bool&)+0x6b) [0x6f6279]
 8: (SimpleMessenger::Pipe::accept()+0xc7d) [0x6ffba5]
 9: (SimpleMessenger::Pipe::reader()+0x32) [0x700de8]
 10: (SimpleMessenger::Pipe::Reader::entry()+0x19) [0x6f4e2d]
 11: (Thread::_entry_func(void*)+0x20) [0x705944]
 12: /lib/libpthread.so.0 [0x7f22962effc7]
 13: (clone()+0x6d) [0x7f22955325ad]
 NOTE: a copy of the executable, or `objdump -rdS <executable>` is needed to interpret this.

15 years agouclient: don't renew caps while mds is recovering
Sage Weil [Wed, 23 Dec 2009 19:48:32 +0000 (11:48 -0800)]
uclient: don't renew caps while mds is recovering

waste of time and memory

15 years agomds: expose projected ctime is client is last versionlock wrlocker
Sage Weil [Wed, 23 Dec 2009 23:54:50 +0000 (15:54 -0800)]
mds: expose projected ctime is client is last versionlock wrlocker

This fixes the problem where the client does a setattr on CTIME, but
doesn't actually see it.

Probably, though, the exposed ctime should be more precise: something like
the most recent projected value that the client is responsible, since we
may have multiple clients projecting pipelined values on the inode.

15 years agotodo
Sage Weil [Wed, 23 Dec 2009 02:32:00 +0000 (18:32 -0800)]
todo

15 years agoosd: prioritized repop acks
Sage Weil [Tue, 22 Dec 2009 22:54:46 +0000 (14:54 -0800)]
osd: prioritized repop acks

This may not make much of a difference, but it won't hurt.

15 years agoqa: don't fail rm of old log
Sage Weil [Tue, 22 Dec 2009 22:50:38 +0000 (14:50 -0800)]
qa: don't fail rm of old log

15 years agoqa: make logging more forgiving
Sage Weil [Tue, 22 Dec 2009 21:45:40 +0000 (13:45 -0800)]
qa: make logging more forgiving

15 years agomsgr: separate incoming queue from local_endpoint
Sage Weil [Wed, 23 Dec 2009 02:03:00 +0000 (18:03 -0800)]
msgr: separate incoming queue from local_endpoint

For now it's bundled into dispatch_queue.

Fixes a race with local_endpoint shutdown and incoming message delivery.

15 years agoosd: put meta_coll const in OSD.{cc,h}
Sage Weil [Wed, 23 Dec 2009 01:50:33 +0000 (17:50 -0800)]
osd: put meta_coll const in OSD.{cc,h}

15 years agofilejournal: delete[] not delete
Sage Weil [Wed, 23 Dec 2009 01:43:18 +0000 (17:43 -0800)]
filejournal: delete[] not delete

15 years agobacktrace: fix #includes for ppc
Sage Weil [Tue, 22 Dec 2009 21:29:07 +0000 (13:29 -0800)]
backtrace: fix #includes for ppc

15 years agohash: use native appropriately sized hash for word, not native host word size
Sage Weil [Tue, 22 Dec 2009 21:26:47 +0000 (13:26 -0800)]
hash: use native appropriately sized hash for word, not native host word size

These shouldn't be used over the wire or on disk, but just in case.

Also, i suspect the 64-bit hash is probably faster than the 32bit one
twice?

15 years agochange LP64 ifdefs to casts for printf formatting
Sage Weil [Tue, 22 Dec 2009 21:25:46 +0000 (13:25 -0800)]
change LP64 ifdefs to casts for printf formatting

15 years agologger: cleanup 64
Sage Weil [Tue, 22 Dec 2009 21:22:56 +0000 (13:22 -0800)]
logger: cleanup 64

15 years agotodo, confs
Sage Weil [Tue, 22 Dec 2009 21:21:33 +0000 (13:21 -0800)]
todo, confs

15 years agomsgr: include features in connection handshake (protocol change)
Sage Weil [Tue, 22 Dec 2009 19:42:29 +0000 (11:42 -0800)]
msgr: include features in connection handshake (protocol change)

15 years agomsgr: put tid in ceph_msg_header (protocol change)
Sage Weil [Tue, 22 Dec 2009 19:41:47 +0000 (11:41 -0800)]
msgr: put tid in ceph_msg_header (protocol change)

15 years agotodo more bugs
Sage Weil [Tue, 22 Dec 2009 00:46:05 +0000 (16:46 -0800)]
todo more bugs

15 years agoassert: leave cruft off top of assert failure backtrace
Sage Weil [Tue, 22 Dec 2009 00:28:27 +0000 (16:28 -0800)]
assert: leave cruft off top of assert failure backtrace

no more
 1: (__ceph_assert_fail(char const*, char const*, int, char const*)+0x34) [0x7ea7ea]

15 years agobacktrace: demangle c++ names
Sage Weil [Tue, 22 Dec 2009 00:26:07 +0000 (16:26 -0800)]
backtrace: demangle c++ names

Finally!

15 years agotodo bugs
Sage Weil [Sun, 20 Dec 2009 05:09:23 +0000 (21:09 -0800)]
todo bugs

15 years agomds: new commands!
Greg Farnum [Sat, 19 Dec 2009 00:13:42 +0000 (16:13 -0800)]
mds: new commands!
issue_caps inodenumber calls locker->issue_caps
try_eval inodenumber mask calls locker->try_eval

15 years agomds: new command kill session xxxx ends that client session
Greg Farnum [Thu, 17 Dec 2009 22:29:13 +0000 (14:29 -0800)]
mds: new command kill session xxxx ends that client session

15 years agocreate a new function end_session to end client sessions
Greg Farnum [Thu, 17 Dec 2009 22:27:26 +0000 (14:27 -0800)]
create a new function end_session to end client sessions

15 years agovstart: allow -o 'some option for config file = whatever'
Sage Weil [Wed, 16 Dec 2009 22:06:01 +0000 (14:06 -0800)]
vstart: allow -o 'some option for config file = whatever'

15 years agoosd: fix PG print log bound check
Sage Weil [Tue, 15 Dec 2009 22:27:56 +0000 (14:27 -0800)]
osd: fix PG print log bound check

15 years agotodo (bugs)
Sage Weil [Thu, 17 Dec 2009 19:32:21 +0000 (11:32 -0800)]
todo (bugs)

15 years agoMake a more flexible 'tell' command to replace injectargs
Greg Farnum [Tue, 15 Dec 2009 23:50:00 +0000 (15:50 -0800)]
Make a more flexible 'tell' command to replace injectargs
osd/mds: make a handle_command function; don't parse all MMonCommands as config

mon: Send injectargs to osd/mds in new format;
accept tells for mds and osd

15 years agotodo (bugs, filestore notes)
Sage Weil [Tue, 15 Dec 2009 22:13:23 +0000 (14:13 -0800)]
todo (bugs, filestore notes)

15 years agoosd: include last_epoch_split in PG::Info::History
Sage Weil [Tue, 15 Dec 2009 22:13:12 +0000 (14:13 -0800)]
osd: include last_epoch_split in PG::Info::History

Also drop backward compat encoding for same_up_since;
overall osd format as reved since then.

15 years agomsgr: don't keep sent messages on lossy connections
Sage Weil [Tue, 15 Dec 2009 22:11:16 +0000 (14:11 -0800)]
msgr: don't keep sent messages on lossy connections

15 years agomds: log client timeouts
Sage Weil [Tue, 15 Dec 2009 22:10:37 +0000 (14:10 -0800)]
mds: log client timeouts

15 years agomds: don't croak on stray ACK
Sage Weil [Mon, 14 Dec 2009 22:14:10 +0000 (14:14 -0800)]
mds: don't croak on stray ACK

This can happen after replay:

- AGREE and update op in journal
- table resends AGREE
- client resends COMMIT
- client gets AGREE, resends COMMIT
- table gets 2 COMMITS, replies with 2 ACKs
- second ACK can be ignored.

15 years agoqa: Copy the output of each test to a log file
Greg Farnum [Tue, 15 Dec 2009 21:23:23 +0000 (13:23 -0800)]
qa: Copy the output of each test to a log file

15 years agoqa: Whoops, never added the runonkclient script
Greg Farnum [Mon, 14 Dec 2009 23:46:06 +0000 (15:46 -0800)]
qa: Whoops, never added the runonkclient script

15 years agoauth: rename mutex so that it doesn't confuse lockdep
Yehuda Sadeh [Mon, 14 Dec 2009 19:27:48 +0000 (11:27 -0800)]
auth: rename mutex so that it doesn't confuse lockdep

15 years agomsgr: maintain per-pipe in_qlen; avoid list.size(); some cleanup
Sage Weil [Fri, 11 Dec 2009 21:40:53 +0000 (13:40 -0800)]
msgr: maintain per-pipe in_qlen; avoid list.size(); some cleanup

15 years agolibrados: fix locking slightly
Sage Weil [Fri, 11 Dec 2009 20:42:43 +0000 (12:42 -0800)]
librados: fix locking slightly

Lock was missing on aio_read.

15 years agoqa: fix runallonce target dir verification
Sage Weil [Fri, 11 Dec 2009 18:19:51 +0000 (10:19 -0800)]
qa: fix runallonce target dir verification

15 years agofilestore: disable USERTRANS for now
Sage Weil [Fri, 11 Dec 2009 18:18:03 +0000 (10:18 -0800)]
filestore: disable USERTRANS for now

15 years agotodo
Sage Weil [Fri, 11 Dec 2009 20:03:29 +0000 (12:03 -0800)]
todo

15 years agolibrados: less noisy
Sage Weil [Fri, 11 Dec 2009 20:03:24 +0000 (12:03 -0800)]
librados: less noisy

15 years agovstart: fix small bash thing
Sage Weil [Fri, 11 Dec 2009 00:51:42 +0000 (16:51 -0800)]
vstart: fix small bash thing

15 years agoosd: drop backlog on recovery completion (when we mark PG_STATE_CLEAN)
Sage Weil [Fri, 11 Dec 2009 16:37:19 +0000 (08:37 -0800)]
osd: drop backlog on recovery completion (when we mark PG_STATE_CLEAN)

15 years agoosd: split pg log too
Sage Weil [Fri, 11 Dec 2009 16:36:53 +0000 (08:36 -0800)]
osd: split pg log too

15 years agoosd: make osd thread pool sizes adjustable
Sage Weil [Fri, 11 Dec 2009 16:34:27 +0000 (08:34 -0800)]
osd: make osd thread pool sizes adjustable

Rename osd_maxthreads to osd_op_threads to be consistent.

15 years agoinit-ceph: use pkill, not killall, so that c$type.$date is also killed
Sage Weil [Fri, 11 Dec 2009 15:52:59 +0000 (07:52 -0800)]
init-ceph: use pkill, not killall, so that c$type.$date is also killed

15 years agomon: fix mon injectargs
Sage Weil [Fri, 11 Dec 2009 05:57:05 +0000 (21:57 -0800)]
mon: fix mon injectargs

If !connection, we know it's local delivery.

15 years agomon: fix null pointer ref on connection
Sage Weil [Fri, 11 Dec 2009 05:50:12 +0000 (21:50 -0800)]
mon: fix null pointer ref on connection

15 years agocrush: fix grammer to parse hash
Sage Weil [Fri, 11 Dec 2009 00:20:06 +0000 (16:20 -0800)]
crush: fix grammer to parse hash

15 years agomsgr: lockdep found some not-actually-possible bad dependencies; they've
Greg Farnum [Thu, 10 Dec 2009 22:57:09 +0000 (14:57 -0800)]
msgr: lockdep found some not-actually-possible bad dependencies; they've
been removed to make it happy

15 years agomsgr: Add at least a little documentation
Greg Farnum [Thu, 10 Dec 2009 22:32:13 +0000 (14:32 -0800)]
msgr: Add at least a little documentation

15 years agomsgr: Remove noisy debug output.
Greg Farnum [Thu, 10 Dec 2009 22:09:22 +0000 (14:09 -0800)]
msgr: Remove noisy debug output.

15 years agomsgr: discard_queue now removes empty pipe queues from endpoint's map
Greg Farnum [Thu, 10 Dec 2009 22:07:25 +0000 (14:07 -0800)]
msgr: discard_queue now removes empty pipe queues from endpoint's map
This prevents bad pointer dereferences

15 years agomsgr: Cover qlen with a Spinlock for faster access and to limit blocking
Greg Farnum [Thu, 10 Dec 2009 19:16:40 +0000 (11:16 -0800)]
msgr: Cover qlen with a Spinlock for faster access and to limit blocking

15 years agomsgr: That unlock's not in the right spot
Greg Farnum [Thu, 10 Dec 2009 03:30:22 +0000 (19:30 -0800)]
msgr: That unlock's not in the right spot

15 years agomsgr: Fix a number of locking and deletion bugs. Appears to work.
Greg Farnum [Thu, 10 Dec 2009 03:10:45 +0000 (19:10 -0800)]
msgr: Fix a number of locking and deletion bugs. Appears to work.

15 years agomsgr: Fix blocking locks and another potential block.
Greg Farnum [Thu, 10 Dec 2009 00:19:27 +0000 (16:19 -0800)]
msgr: Fix blocking locks and another potential block.

15 years agomsgr: rename locks for clarity; move some around to prevent blocking.
Greg Farnum [Wed, 9 Dec 2009 23:12:00 +0000 (15:12 -0800)]
msgr: rename locks for clarity; move some around to prevent blocking.

15 years agomsgr: implement round-robin pipe message delivery.
Greg Farnum [Wed, 9 Dec 2009 03:52:20 +0000 (19:52 -0800)]
msgr: implement round-robin pipe message delivery.

15 years agomsgr: Remove Endpoint::dispatch_queue and queue_message;
Greg Farnum [Wed, 9 Dec 2009 00:17:57 +0000 (16:17 -0800)]
msgr: Remove Endpoint::dispatch_queue and queue_message;
set up local pipe and necessary routing code.

Also, fix potential locking issue by pushing message onto
queue before pushing queue to round-robin

15 years agomsgr: Add queues, etc needed for message delivery by round-robin pipe
Greg Farnum [Tue, 8 Dec 2009 01:58:33 +0000 (17:58 -0800)]
msgr: Add queues, etc needed for message delivery by round-robin pipe

15 years agomsgr: rename pipe's outbound message queue
Greg Farnum [Mon, 7 Dec 2009 22:30:54 +0000 (14:30 -0800)]
msgr: rename pipe's outbound message queue

15 years agoconfig: parse 'lockdep' option in .conf
Sage Weil [Thu, 10 Dec 2009 21:01:35 +0000 (13:01 -0800)]
config: parse 'lockdep' option in .conf

15 years agoceph.conf.sepia: more osds, lockdep
Sage Weil [Thu, 10 Dec 2009 21:01:27 +0000 (13:01 -0800)]
ceph.conf.sepia: more osds, lockdep

15 years agomon: log error, don't abort, on stray pg_stat notify
Sage Weil [Thu, 10 Dec 2009 18:03:44 +0000 (10:03 -0800)]
mon: log error, don't abort, on stray pg_stat notify

15 years agoosd: set pg_stat.created
Sage Weil [Thu, 10 Dec 2009 17:56:54 +0000 (09:56 -0800)]
osd: set pg_stat.created

15 years agoosd: respond to query on uncreated pg with empty info; do not create
Sage Weil [Thu, 10 Dec 2009 17:46:15 +0000 (09:46 -0800)]
osd: respond to query on uncreated pg with empty info; do not create

We should never create a PG* for an uncreated pg except through the mkpg
path.

15 years agomon: stop trying to create localized pgs >= max osd|device
Sage Weil [Thu, 10 Dec 2009 16:28:31 +0000 (08:28 -0800)]
mon: stop trying to create localized pgs >= max osd|device

If lpg_num == lpgp_num, and the preferred osd no longer exists, there is
no guarantee that the parent and child pgs will land on the same osd.  This
happens when we are creating pgs for a new osd, they don't finish getting
created, and then we destroy that OSD again.

Care should be taken to avoid expanding the OSD count _and_ lpg_num
simultaneously...