]> git.apps.os.sepia.ceph.com Git - ceph.git/log
ceph.git
16 years agoosd: print ops args correctly
Sage Weil [Fri, 30 Jan 2009 23:00:45 +0000 (15:00 -0800)]
osd: print ops args correctly

16 years agomds: use generic helper when removing a cap
Sage Weil [Fri, 30 Jan 2009 23:00:34 +0000 (15:00 -0800)]
mds: use generic helper when removing a cap

This way we make sure to do all the right checks on the locks,
mds_caps, and to call eval_stray if the inode is unlinked.

16 years agomds: fix try_remove_unlinked_dn
Sage Weil [Fri, 30 Jan 2009 22:59:24 +0000 (14:59 -0800)]
mds: fix try_remove_unlinked_dn

Clear dentry NEW flag when we commit it.. pretty sure that was the
old bug.  And we can't cheat the directory commit because the fnode
also gets updated.

16 years agomds: fix purge of 0-length inode
Sage Weil [Fri, 30 Jan 2009 21:51:49 +0000 (13:51 -0800)]
mds: fix purge of 0-length inode

16 years agomds: remove unused inode purge fields
Sage Weil [Fri, 30 Jan 2009 21:15:20 +0000 (13:15 -0800)]
mds: remove unused inode purge fields

16 years agoceph_init: fix initalization
Yehuda Sadeh [Fri, 30 Jan 2009 23:56:36 +0000 (15:56 -0800)]
ceph_init: fix initalization

16 years agoceph_init: added missing file
Yehuda Sadeh [Fri, 30 Jan 2009 23:48:56 +0000 (15:48 -0800)]
ceph_init: added missing file

16 years agoceph_init: initial version
Yehuda Sadeh [Fri, 30 Jan 2009 23:47:59 +0000 (15:47 -0800)]
ceph_init: initial version

16 years agocrun: restart only processes that should have dumped core
Yehuda Sadeh [Fri, 30 Jan 2009 20:45:31 +0000 (12:45 -0800)]
crun: restart only processes that should have dumped core

16 years agotodo
Sage Weil [Fri, 30 Jan 2009 19:43:53 +0000 (11:43 -0800)]
todo

16 years agoosd: put full original reqid in osd_op_reply
Sage Weil [Fri, 30 Jan 2009 19:40:40 +0000 (11:40 -0800)]
osd: put full original reqid in osd_op_reply

16 years agomds: typos
Sage Weil [Fri, 30 Jan 2009 19:36:37 +0000 (11:36 -0800)]
mds: typos

16 years agokclient: don't retain caps if inode is closed and nlink == 0
Sage Weil [Fri, 30 Jan 2009 19:36:28 +0000 (11:36 -0800)]
kclient: don't retain caps if inode is closed and nlink == 0

16 years agomds: rip out old purge_inode crap
Sage Weil [Fri, 30 Jan 2009 19:22:40 +0000 (11:22 -0800)]
mds: rip out old purge_inode crap

16 years agomds: simplify purge_stray
Sage Weil [Fri, 30 Jan 2009 18:41:12 +0000 (10:41 -0800)]
mds: simplify purge_stray

Just delete objects and then journal completion.  No need to do
the funky purge_inode double commit crap, since we will re-eval
strays later any (and do it all again if needed).

16 years agomds: clean up opent a bit
Sage Weil [Fri, 30 Jan 2009 18:08:44 +0000 (10:08 -0800)]
mds: clean up opent a bit

16 years agomds: drop _do_open helper
Sage Weil [Fri, 30 Jan 2009 18:07:06 +0000 (10:07 -0800)]
mds: drop _do_open helper

only one caller

16 years agomds: open O_TRUNC implemented.. but not used by kclient :/
Sage Weil [Fri, 30 Jan 2009 17:56:13 +0000 (09:56 -0800)]
mds: open O_TRUNC implemented.. but not used by kclient :/

16 years agomds: implement async truncate (trimtrunc); fix log segment trim vs purge bug
Sage Weil [Fri, 30 Jan 2009 00:53:10 +0000 (16:53 -0800)]
mds: implement async truncate (trimtrunc); fix log segment trim vs purge bug

Logsegs were allowed to trim even with outstanding purges.. fix.

16 years agoosd: fix truncate ops; add TRIMTRUNC
Sage Weil [Thu, 29 Jan 2009 22:47:46 +0000 (14:47 -0800)]
osd: fix truncate ops; add TRIMTRUNC

16 years agomds: truncate changes. not complete.
Sage Weil [Thu, 29 Jan 2009 22:47:31 +0000 (14:47 -0800)]
mds: truncate changes.  not complete.

We don't clean up after we finish our truncate.

16 years agokclient: fix and clean up osd request generation
Sage Weil [Thu, 29 Jan 2009 22:47:06 +0000 (14:47 -0800)]
kclient: fix and clean up osd request generation

Add trunc op correctly.  Clean things up by combining request
message and request struct creation.

16 years agokclient: include old size in truncate
Sage Weil [Thu, 29 Jan 2009 22:46:35 +0000 (14:46 -0800)]
kclient: include old size in truncate

16 years agokclient: truncate bits
Sage Weil [Thu, 29 Jan 2009 22:46:23 +0000 (14:46 -0800)]
kclient: truncate bits

Queue vmtruncate if we see size decrease in fill_file_info (not
just on CAP_OP_TRUNC).

16 years agokclient: include truncate osd_ops when needed
Sage Weil [Thu, 29 Jan 2009 19:22:15 +0000 (11:22 -0800)]
kclient: include truncate osd_ops when needed

16 years agoosd: MASKTRUNC and SETTRUNC osd ops
Sage Weil [Thu, 29 Jan 2009 18:58:50 +0000 (10:58 -0800)]
osd: MASKTRUNC and SETTRUNC osd ops

16 years agomds: add truncate_size
Sage Weil [Sat, 24 Jan 2009 05:22:57 +0000 (21:22 -0800)]
mds: add truncate_size

16 years agorenamed common.sh to ceph_common.sh
Yehuda Sadeh [Fri, 30 Jan 2009 19:41:06 +0000 (11:41 -0800)]
renamed common.sh to ceph_common.sh

16 years agorename stop.sh to ceph_stop, added to Makefile.am
Yehuda Sadeh [Fri, 30 Jan 2009 19:23:52 +0000 (11:23 -0800)]
rename stop.sh to ceph_stop, added to Makefile.am

16 years agocephd: update Makefile.am
Yehuda Sadeh [Fri, 30 Jan 2009 19:12:19 +0000 (11:12 -0800)]
cephd: update Makefile.am

16 years agomon: only propose new osdmap once at end of mdsmonitor tick()
Sage Weil [Fri, 30 Jan 2009 19:13:23 +0000 (11:13 -0800)]
mon: only propose new osdmap once at end of mdsmonitor tick()

16 years agoceph.sh renamed to cephd
Yehuda Sadeh [Fri, 30 Jan 2009 19:01:41 +0000 (11:01 -0800)]
ceph.sh renamed to cephd

16 years agomsgr: fix WAIT connection race handling
Sage Weil [Fri, 30 Jan 2009 19:00:43 +0000 (11:00 -0800)]
msgr: fix WAIT connection race handling

Not sure why it wasn't setting STATE_WAIT before.  Works now.

16 years agoceph.sh: some changes, fixes
Yehuda Sadeh [Fri, 30 Jan 2009 18:37:51 +0000 (10:37 -0800)]
ceph.sh: some changes, fixes

16 years agoceph.sh: some fixes
Yehuda Sadeh [Fri, 30 Jan 2009 00:20:46 +0000 (16:20 -0800)]
ceph.sh: some fixes

16 years agoceph.sh: try to locate cconf
Yehuda Sadeh [Thu, 29 Jan 2009 23:30:17 +0000 (15:30 -0800)]
ceph.sh: try to locate cconf

16 years agoceph.sh: enable script
Yehuda Sadeh [Thu, 29 Jan 2009 23:24:16 +0000 (15:24 -0800)]
ceph.sh: enable script

16 years agoceph.sh: able to specify start/stop/restart, specific modules
Yehuda Sadeh [Thu, 29 Jan 2009 23:22:52 +0000 (15:22 -0800)]
ceph.sh: able to specify start/stop/restart, specific modules

16 years agoceph.sh: add missing file
Yehuda Sadeh [Thu, 29 Jan 2009 23:02:34 +0000 (15:02 -0800)]
ceph.sh: add missing file

16 years agoceph.sh: can specify remote modules
Yehuda Sadeh [Thu, 29 Jan 2009 23:01:25 +0000 (15:01 -0800)]
ceph.sh: can specify remote modules

16 years agoceph.sh: startup script that uses conf file
Yehuda Sadeh [Thu, 29 Jan 2009 21:55:33 +0000 (13:55 -0800)]
ceph.sh: startup script that uses conf file

16 years agoosd: rev internal osd protocol
Sage Weil [Thu, 29 Jan 2009 18:16:12 +0000 (10:16 -0800)]
osd: rev internal osd protocol

16 years agoosd: push wrlocks into object_info_t
Sage Weil [Wed, 28 Jan 2009 21:06:05 +0000 (13:06 -0800)]
osd: push wrlocks into object_info_t

16 years agoosd: remove inc_lock
Sage Weil [Wed, 28 Jan 2009 19:10:21 +0000 (11:10 -0800)]
osd: remove inc_lock

16 years agoosd: combine all internal object state into single '_' object_info_t xattr
Sage Weil [Wed, 28 Jan 2009 19:08:14 +0000 (11:08 -0800)]
osd: combine all internal object state into single '_' object_info_t xattr

16 years agoosd: replace version attr with oi (object_info_t) (disk format change)
Sage Weil [Tue, 27 Jan 2009 19:20:37 +0000 (11:20 -0800)]
osd: replace version attr with oi (object_info_t) (disk format change)

Includes last reqid, mtime.  Makes backlog log entries include
a real reqid, which makes the client_reqs IndexedLog map meaningful.
Otherwise, the first backlog item we 'unindex' means the remaining
ops with reqid unknown0.0:0 aren't in the map.

This revs the osd ondisk format.

16 years agoMerge branch 'unstable' of ssh://ceph.newdream.net/git/ceph into unstable
Yehuda Sadeh [Thu, 29 Jan 2009 00:35:54 +0000 (16:35 -0800)]
Merge branch 'unstable' of ssh://ceph.newdream.net/git/ceph into unstable

16 years agocconf: able to specify multiple sections
Yehuda Sadeh [Thu, 29 Jan 2009 00:35:39 +0000 (16:35 -0800)]
cconf: able to specify multiple sections

16 years agodstart: --ramjournal option
Sage Weil [Thu, 29 Jan 2009 00:11:07 +0000 (16:11 -0800)]
dstart: --ramjournal option

16 years agoscripts: osd latency check script
Sage Weil [Thu, 29 Jan 2009 00:10:57 +0000 (16:10 -0800)]
scripts: osd latency check script

16 years agokclient: start writeback without blocking
Sage Weil [Wed, 28 Jan 2009 23:52:58 +0000 (15:52 -0800)]
kclient: start writeback without blocking

We just want to make sure writeback starts; we don't need/want
to wait for it to complete in the writeback thread.

16 years agocconf: set default conf filename
Yehuda Sadeh [Wed, 28 Jan 2009 18:51:18 +0000 (10:51 -0800)]
cconf: set default conf filename

16 years agocconf: added utility to read configuration
Yehuda Sadeh [Wed, 28 Jan 2009 18:49:37 +0000 (10:49 -0800)]
cconf: added utility to read configuration

16 years agomds: fix open file rejounal in try_to_expire
Sage Weil [Wed, 28 Jan 2009 17:25:37 +0000 (09:25 -0800)]
mds: fix open file rejounal in try_to_expire

We were invalidating the iterator.

16 years agokclient: fix nocrc behavior
Sage Weil [Wed, 28 Jan 2009 17:32:16 +0000 (09:32 -0800)]
kclient: fix nocrc behavior

Only disable data crc.  Don't check crc if either we or the peer
didn't calculate it.

16 years agomds: shorten mds beacon grace to 15 seconds
Sage Weil [Wed, 28 Jan 2009 00:29:10 +0000 (16:29 -0800)]
mds: shorten mds beacon grace to 15 seconds

16 years agotodo
Sage Weil [Wed, 28 Jan 2009 00:18:06 +0000 (16:18 -0800)]
todo

16 years agovstart: fix btrfs test
Sage Weil [Tue, 27 Jan 2009 23:20:49 +0000 (15:20 -0800)]
vstart: fix btrfs test

16 years agomsgr: make nocrc only apply to data payload; and set flag if no crc
Sage Weil [Tue, 27 Jan 2009 22:56:25 +0000 (14:56 -0800)]
msgr: make nocrc only apply to data payload; and set flag if no crc

16 years agotodo
Sage Weil [Tue, 27 Jan 2009 22:42:31 +0000 (14:42 -0800)]
todo

16 years agomon: clean mon monmap cons
Sage Weil [Tue, 27 Jan 2009 22:42:27 +0000 (14:42 -0800)]
mon: clean mon monmap cons

16 years agodstart: no preload gprof helper
Sage Weil [Tue, 27 Jan 2009 22:42:17 +0000 (14:42 -0800)]
dstart: no preload gprof helper

16 years agoconfig: avoid wonky times in g_config
Sage Weil [Tue, 27 Jan 2009 22:15:00 +0000 (14:15 -0800)]
config: avoid wonky times in g_config

16 years agomon: disable pgmap_dumps
Sage Weil [Tue, 27 Jan 2009 22:19:46 +0000 (14:19 -0800)]
mon: disable pgmap_dumps

This aren't trimmed!

16 years agomds: optimize EMetaBlob::fullbit, remotebit, nullbit encoding
Sage Weil [Tue, 27 Jan 2009 21:51:51 +0000 (13:51 -0800)]
mds: optimize EMetaBlob::fullbit, remotebit, nullbit encoding

Encode in constructor to avoid intermediate representation during
journaling.

16 years agobuffer: optimize append()
Sage Weil [Tue, 27 Jan 2009 21:23:57 +0000 (13:23 -0800)]
buffer: optimize append()

Avoid repeatedly dereferencing list tail.

16 years agokclient: fix osd timeout merge error
Sage Weil [Tue, 27 Jan 2009 18:46:18 +0000 (10:46 -0800)]
kclient: fix osd timeout merge error

16 years agodstart: less debug output
Sage Weil [Tue, 27 Jan 2009 18:42:55 +0000 (10:42 -0800)]
dstart: less debug output

16 years agoosd: throttle sub_ops too
Sage Weil [Tue, 27 Jan 2009 18:27:39 +0000 (10:27 -0800)]
osd: throttle sub_ops too

16 years agoosd: adjust pending_ops when requeuing opq
Sage Weil [Tue, 27 Jan 2009 17:48:37 +0000 (09:48 -0800)]
osd: adjust pending_ops when requeuing opq

16 years agomds: show which segment trimming is blocked on
Sage Weil [Tue, 27 Jan 2009 00:30:11 +0000 (16:30 -0800)]
mds: show which segment trimming is blocked on

16 years agoosd: locally requeue on repeer if still primary
Sage Weil [Tue, 27 Jan 2009 18:40:11 +0000 (10:40 -0800)]
osd: locally requeue on repeer if still primary

kclient osd_client and Objecter need only resend when primary
changes.

16 years agokclient: fix RETRY_SESSION seq value
Sage Weil [Mon, 26 Jan 2009 20:13:26 +0000 (12:13 -0800)]
kclient: fix RETRY_SESSION seq value

Should be existing session's seq, not new sessions.  Error causes
SimpleMessenger to throw an assertion

msg/SimpleMessenger.cc:1283: FAILED assert(reply.connect_seq > connect_seq)

16 years agokclient: only send one ping per laggy osd
Sage Weil [Tue, 27 Jan 2009 04:57:06 +0000 (20:57 -0800)]
kclient: only send one ping per laggy osd

16 years agomds: include xattr updates in cap messages; accept xattr cap updates
Sage Weil [Tue, 27 Jan 2009 04:21:17 +0000 (20:21 -0800)]
mds: include xattr updates in cap messages; accept xattr cap updates

16 years agomds: adjust some debug levels
Sage Weil [Tue, 27 Jan 2009 03:56:40 +0000 (19:56 -0800)]
mds: adjust some debug levels

16 years agotodo
Sage Weil [Sat, 24 Jan 2009 05:23:38 +0000 (21:23 -0800)]
todo

16 years agoosd: drop osd_lock while pausing/requeueing opq
Sage Weil [Tue, 27 Jan 2009 00:21:36 +0000 (16:21 -0800)]
osd: drop osd_lock while pausing/requeueing opq

An in-progress op may need to take osd_lock.

16 years agoosd: fix merge_old_entry assertion
Sage Weil [Tue, 27 Jan 2009 00:05:17 +0000 (16:05 -0800)]
osd: fix merge_old_entry assertion

Object may not be missing if it was later deleted.

16 years agoosd: fix up_thru_pending conditional
Sage Weil [Mon, 26 Jan 2009 23:25:56 +0000 (15:25 -0800)]
osd: fix up_thru_pending conditional

up_thru_pending only meaningful if non-zero.

16 years agoosd: requeue instead of draining op_queue on map update
Sage Weil [Mon, 26 Jan 2009 23:24:04 +0000 (15:24 -0800)]
osd: requeue instead of draining op_queue on map update

Also clean up waiter helpers.

16 years agoosd: keep peer_info.last_{update,complete} up to date in case we get later stray
Sage Weil [Mon, 26 Jan 2009 21:50:53 +0000 (13:50 -0800)]
osd: keep peer_info.last_{update,complete} up to date in case we get later stray

We need to be able to call peer() again after going active and get
meaningful results.

16 years agoosd: bit more verbose in oldest_update calculation
Sage Weil [Mon, 26 Jan 2009 20:42:36 +0000 (12:42 -0800)]
osd: bit more verbose in oldest_update calculation

16 years agoosd: adjust generate_backlog assertions
Sage Weil [Mon, 26 Jan 2009 20:42:14 +0000 (12:42 -0800)]
osd: adjust generate_backlog assertions

We should never generate a backlog while active.

16 years agomds: clean up EUpdate output
Sage Weil [Mon, 26 Jan 2009 20:06:12 +0000 (12:06 -0800)]
mds: clean up EUpdate output

16 years agomds: do not trim log during recovery
Sage Weil [Mon, 26 Jan 2009 20:06:03 +0000 (12:06 -0800)]
mds: do not trim log during recovery

16 years agoosd: only check replicas in recover_replicas
Sage Weil [Mon, 26 Jan 2009 19:51:43 +0000 (11:51 -0800)]
osd: only check replicas in recover_replicas

16 years agoosd: clean up activate, peer recovery debug output
Sage Weil [Mon, 26 Jan 2009 19:22:25 +0000 (11:22 -0800)]
osd: clean up activate, peer recovery debug output

16 years agomessenger: added a no-crc option
Yehuda Sadeh [Mon, 26 Jan 2009 20:47:13 +0000 (12:47 -0800)]
messenger: added a no-crc option

16 years agoosd: fix pg log trimming
Sage Weil [Mon, 26 Jan 2009 19:23:25 +0000 (11:23 -0800)]
osd: fix pg log trimming

Log entries rarely ever fall on block boundaries.  Behave
accordingly.

Also, decide whether to trim on primary only, at a high level in
op_modify.

16 years agopaxos: tolerate commits for older states
Sage Weil [Mon, 26 Jan 2009 19:06:42 +0000 (11:06 -0800)]
paxos: tolerate commits for older states

We may have races if paxos recovery restarts.

16 years agoosd: fix repop cleanup in on_change, _shutdown
Sage Weil [Mon, 26 Jan 2009 18:14:07 +0000 (10:14 -0800)]
osd: fix repop cleanup in on_change, _shutdown

16 years agoceph: warn when we get message with wrong fsid
Sage Weil [Mon, 26 Jan 2009 18:13:44 +0000 (10:13 -0800)]
ceph: warn when we get message with wrong fsid

16 years agotodos
Sage Weil [Sat, 24 Jan 2009 00:47:18 +0000 (16:47 -0800)]
todos

16 years agopaxos: ignore LAST message if pn is old
Sage Weil [Fri, 23 Jan 2009 23:21:04 +0000 (15:21 -0800)]
paxos: ignore LAST message if pn is old

And old PN means that it's from a prior COLLECT attempt.

16 years agomds: increase ino prealloc limit to 1000
Sage Weil [Sat, 24 Jan 2009 00:39:46 +0000 (16:39 -0800)]
mds: increase ino prealloc limit to 1000

Ideally this should self-tune, somehow...

16 years agoosd: remove RepGather::put() debug print
Sage Weil [Sat, 24 Jan 2009 00:39:26 +0000 (16:39 -0800)]
osd: remove RepGather::put() debug print

16 years agocrush: remote list builder printf
Sage Weil [Sat, 24 Jan 2009 00:39:03 +0000 (16:39 -0800)]
crush: remote list builder printf

16 years agodstart: mount btrfs with new options
Sage Weil [Fri, 23 Jan 2009 20:53:18 +0000 (12:53 -0800)]
dstart: mount btrfs with new options

16 years agokclient: statfs_request_tree alloc context is NOFS, not ATOMIC
Sage Weil [Sat, 24 Jan 2009 00:43:09 +0000 (16:43 -0800)]
kclient: statfs_request_tree alloc context is NOFS, not ATOMIC