]> git.apps.os.sepia.ceph.com Git - ceph.git/log
ceph.git
13 years agofix json escaping
Yehuda Sadeh [Tue, 30 Aug 2011 00:01:41 +0000 (17:01 -0700)]
fix json escaping

13 years agoclient: debug link/unlink and dn_set parent relationship
Sage Weil [Mon, 29 Aug 2011 22:53:57 +0000 (15:53 -0700)]
client: debug link/unlink and dn_set parent relationship

Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
13 years agoheartbeatmap: fix reset_timeout with mixed-used threads
Sage Weil [Mon, 29 Aug 2011 22:02:37 +0000 (15:02 -0700)]
heartbeatmap: fix reset_timeout with mixed-used threads

If you have a ThreadPool used by multiple WorkQueues, and some of them are
setting a suicide timeout, we need to clear it when a suicide timeout is
not set.

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agorados tool: df dumps formatted data
Yehuda Sadeh [Mon, 29 Aug 2011 22:04:59 +0000 (15:04 -0700)]
rados tool: df dumps formatted data

13 years agorgw: log show dumps some information about bucket owner
Yehuda Sadeh [Mon, 29 Aug 2011 21:19:43 +0000 (14:19 -0700)]
rgw: log show dumps some information about bucket owner

13 years agobandaid for gitbuider
Samuel Just [Mon, 29 Aug 2011 20:06:26 +0000 (13:06 -0700)]
bandaid for gitbuider

13 years agoMerge remote branch 'vapre/wip-pgid'
Sage Weil [Mon, 29 Aug 2011 20:39:12 +0000 (13:39 -0700)]
Merge remote branch 'vapre/wip-pgid'

13 years agoosd: remove dead lookup_lock_pg()
Sage Weil [Mon, 29 Aug 2011 20:39:10 +0000 (13:39 -0700)]
osd: remove dead lookup_lock_pg()

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agoosd: set suicide timeouts on some workqueues
Sage Weil [Mon, 29 Aug 2011 18:54:21 +0000 (11:54 -0700)]
osd: set suicide timeouts on some workqueues

OpWQ: timeout * 10
RecoveryWQ: this does no io; it if stalls we're probably stuck in an
  infinite loop.  timeout * 10.
ScrubFinalizeWQ: this is cpu only. we're probably stuck in a loop, or
  swapping.  timeout * 10.

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agomon: health not ok when up < in osds
Sage Weil [Mon, 29 Aug 2011 18:41:24 +0000 (11:41 -0700)]
mon: health not ok when up < in osds

We were warning if there were any not up or in osds.  Instead, warn if
there are any osds that are in but not up.  That means if a node fails
and successfully marks the node out and retracts onto remaining nodes, the
ceph cluster is healthy again.

Presumably the fact that the nodes failed should raise other alerts,
because those specific daemons/nodes are not healthy.

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agofilestore: add suicide grace to workqueue
Sage Weil [Mon, 29 Aug 2011 17:08:08 +0000 (10:08 -0700)]
filestore: add suicide grace to workqueue

The main thing we are worried about here are kernel BUGs that kill off our
threads.

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agoheartbeatmap: add suicide grace
Sage Weil [Mon, 29 Aug 2011 17:04:30 +0000 (10:04 -0700)]
heartbeatmap: add suicide grace

Add a second suicide grace period that will make us kill ourselves if
we are sufficiently catatonic.

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agorgw: don't check for time skew when Expires was set
Yehuda Sadeh [Mon, 29 Aug 2011 17:57:19 +0000 (10:57 -0700)]
rgw: don't check for time skew when Expires was set

13 years agoclient: simplify Client::remove_cap() args
Sage Weil [Mon, 29 Aug 2011 16:27:13 +0000 (09:27 -0700)]
client: simplify Client::remove_cap() args

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agoclient: clean up Client::remove_session_caps() args
Sage Weil [Mon, 29 Aug 2011 16:21:33 +0000 (09:21 -0700)]
client: clean up Client::remove_session_caps() args

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agoclient: don't falsely populate mds_sessions on bad MClientSession msg
Sage Weil [Mon, 29 Aug 2011 16:20:53 +0000 (09:20 -0700)]
client: don't falsely populate mds_sessions on bad MClientSession msg

If from isn't in the map, we don't want to add it via operator[].

Fixes: #1367
Signed-off-by: Sage Weil <sage@newdream.net>
13 years agotestlibrbd: remove useless sizeofs
Sage Weil [Mon, 29 Aug 2011 15:59:46 +0000 (08:59 -0700)]
testlibrbd: remove useless sizeofs

Should be sizeof(*result), if anything.

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agotestlibrbd: fix assert side effect
Sage Weil [Mon, 29 Aug 2011 15:55:53 +0000 (08:55 -0700)]
testlibrbd: fix assert side effect

Coverity cid 1, 2

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agoFileJournal: Also classify kernel major version 3 as safe with disk-cache on
Wido den Hollander [Sat, 27 Aug 2011 20:37:05 +0000 (22:37 +0200)]
FileJournal: Also classify kernel major version 3 as safe with disk-cache on

Signed-off-by: Wido den Hollander <wido@widodh.nl>
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
13 years agoosd: decode pg_t in old PG::Info
Sage Weil [Sun, 28 Aug 2011 03:45:15 +0000 (20:45 -0700)]
osd: decode pg_t in old PG::Info

Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
13 years agoosd: requrie PGID64 of mons, other osds
Sage Weil [Sat, 27 Aug 2011 16:57:13 +0000 (09:57 -0700)]
osd: requrie PGID64 of mons, other osds

Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
13 years agomds: require PGID64 of mons
Sage Weil [Sat, 27 Aug 2011 16:56:45 +0000 (09:56 -0700)]
mds: require PGID64 of mons

Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
13 years agomon: require PGID64 of osds and other mons
Sage Weil [Sat, 27 Aug 2011 16:56:07 +0000 (09:56 -0700)]
mon: require PGID64 of osds and other mons

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agopgmap: decode old pg_t
Sage Weil [Fri, 26 Aug 2011 19:31:52 +0000 (12:31 -0700)]
pgmap: decode old pg_t

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agomdsmap: int64_t pool ids
Sage Weil [Fri, 26 Aug 2011 18:01:41 +0000 (11:01 -0700)]
mdsmap: int64_t pool ids

Encode old-format maps for clients without PGID64 feature bit.

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agoosd: add 64-bit pools to ObjectLocator
Sage Weil [Fri, 26 Aug 2011 17:03:04 +0000 (10:03 -0700)]
osd: add 64-bit pools to ObjectLocator

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agoosdmap: encode/decode old + new version
Sage Weil [Fri, 26 Aug 2011 16:55:20 +0000 (09:55 -0700)]
osdmap: encode/decode old + new version

In MOSDMap, reencode map payloads using old format if the target doesn't
have the PGID64 feature bit.

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agoosd: decode old pg_stat_t
Sage Weil [Thu, 25 Aug 2011 21:08:52 +0000 (14:08 -0700)]
osd: decode old pg_stat_t

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agoosd: make MOSDOp[Reply] encoding backwards compatible
Sage Weil [Thu, 25 Aug 2011 20:59:42 +0000 (13:59 -0700)]
osd: make MOSDOp[Reply] encoding backwards compatible

Move away from struct-based encoding of MOSDOpReply while we're at it.

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agoosd: add PGID64 feature bit
Sage Weil [Thu, 25 Aug 2011 20:52:30 +0000 (13:52 -0700)]
osd: add PGID64 feature bit

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agoosd: int pool -> int64_t pool
Sage Weil [Thu, 25 Aug 2011 20:51:29 +0000 (13:51 -0700)]
osd: int pool -> int64_t pool

Internal interfaces.

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agoosd: redo pg_t encoding with 64-bit pool id
Sage Weil [Thu, 25 Aug 2011 20:30:49 +0000 (13:30 -0700)]
osd: redo pg_t encoding with 64-bit pool id

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agoosd: fix pg_pool_t::get_pg_num return types
Sage Weil [Thu, 25 Aug 2011 20:28:27 +0000 (13:28 -0700)]
osd: fix pg_pool_t::get_pg_num return types

Unsigned!

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agoceph_context: whitespace
Sage Weil [Fri, 26 Aug 2011 04:29:52 +0000 (21:29 -0700)]
ceph_context: whitespace

Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
13 years agoMerge commit 'v0.34'
Sage Weil [Sat, 27 Aug 2011 16:27:41 +0000 (09:27 -0700)]
Merge commit 'v0.34'

13 years agov0.34 v0.34
Sage Weil [Sat, 27 Aug 2011 04:48:47 +0000 (21:48 -0700)]
v0.34

13 years agorgw: use formatter->dump_string() instead of dump_format()
Yehuda Sadeh [Fri, 26 Aug 2011 23:54:05 +0000 (16:54 -0700)]
rgw: use formatter->dump_string() instead of dump_format()

13 years agorgw: get bucket info only if bucket name is not empty
Yehuda Sadeh [Fri, 26 Aug 2011 21:10:25 +0000 (14:10 -0700)]
rgw: get bucket info only if bucket name is not empty

13 years agoMakefile.am: install coverage files for libtool outputs as well
Josh Durgin [Thu, 11 Aug 2011 19:01:15 +0000 (12:01 -0700)]
Makefile.am: install coverage files for libtool outputs as well

Signed-off-by: Josh Durgin <josh.durgin@dreamhost.com>
13 years agolibrados: watch request should wait for safe response
Yehuda Sadeh [Fri, 26 Aug 2011 18:43:47 +0000 (11:43 -0700)]
librados: watch request should wait for safe response

13 years agoobjectcacher: only want for commit
Sage Weil [Fri, 26 Aug 2011 16:47:03 +0000 (09:47 -0700)]
objectcacher: only want for commit

There was some old, weird stuff going on here where we would wait for the
ACK and COMMIT separately.  This is just wrong.  Writeback does not
complete until the data is committed on disk.

Simplify by waiting only for commit, removing all the 'ack' code, and
going back to a single callback (flush_set).

I didn't notice this for 05063867e2a54176ffc9bbc73391f52766ab403f; both of
these cleanups are needed to fix this.

Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
13 years agomds: fix shadow
Sage Weil [Fri, 26 Aug 2011 16:10:21 +0000 (09:10 -0700)]
mds: fix shadow

Coverity cid 29

Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
13 years agofix utime.h #include recursion
Sage Weil [Fri, 26 Aug 2011 16:09:45 +0000 (09:09 -0700)]
fix utime.h #include recursion

Coverity cid 11, 13

Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
13 years agocrush: comment fall-thru
Sage Weil [Fri, 26 Aug 2011 16:00:08 +0000 (09:00 -0700)]
crush: comment fall-thru

Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
13 years agomds: remove unused Mutation in do_rmdir_rollback()
Sage Weil [Thu, 25 Aug 2011 19:28:32 +0000 (12:28 -0700)]
mds: remove unused Mutation in do_rmdir_rollback()

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agoworkunits: make file_layout test setting only one parameter
Greg Farnum [Thu, 25 Aug 2011 19:40:55 +0000 (12:40 -0700)]
workunits: make file_layout test setting only one parameter

Signed-off-by: Greg Farnum <gregory.farnum@dreamhost.com>
13 years agocephfs: use strtol instead of atoi; handle 0 properly
Greg Farnum [Thu, 25 Aug 2011 19:22:02 +0000 (12:22 -0700)]
cephfs: use strtol instead of atoi; handle 0 properly

Besides being generally better, this means we can accept pool 0
as the pool to store stuff in.

Signed-off-by: Greg Farnum <gregory.farnum@dreamhost.com>
13 years agocephfs: fix the stripe_count documentation.
Greg Farnum [Thu, 25 Aug 2011 16:24:58 +0000 (09:24 -0700)]
cephfs: fix the stripe_count documentation.

Whoops!

Signed-off-by: Greg Farnum <gregory.farnum@dreamhost.com>
13 years agomds: server: should apply new layout settings on top of old layout
Greg Farnum [Thu, 25 Aug 2011 18:28:42 +0000 (11:28 -0700)]
mds: server: should apply new layout settings on top of old layout

This way, the MDS can handle updates of some values without needing
the user to specify the entire layout (ie, they can just switch pools).
This brings the behavior more in line with setting the dir layout.

Signed-off-by: Greg Farnum <gregory.farnum@dreamhost.com>
13 years agomds: allow setting all options on a file layout.
Greg Farnum [Thu, 25 Aug 2011 16:49:38 +0000 (09:49 -0700)]
mds: allow setting all options on a file layout.

Signed-off-by: Greg Farnum <gregory.farnum@dreamhost.com>
13 years agoobjectcacher: remove uncommitted xlist<>; fix broken purge() cleanup
Sage Weil [Thu, 25 Aug 2011 19:37:18 +0000 (12:37 -0700)]
objectcacher: remove uncommitted xlist<>; fix broken purge() cleanup

There was a problem where:

 - we would dirty some buffers on an object
   - bump dirty_tx count
 - flush()
   - this adds the Object to ObjectSet::uncommitted
 - truncate
   - client clears FILE_BUFFER cap_ref
   - Object::purge()
     - clear dirty_tx count
 - client puts last inode
   - Object::uncommitted is not empty in ~ObjectSet

(This was triggered after several runs of workunits/suites/blogbensh.sh
on sepia.)

It turns out the uncommitted xlist<> is pretty useless, though: the same
information is captured in the dirty_tx counter.  We add a separate
counter to the Object itself (for the benefit of Object::can_close()).

We also clean up Object::purge() to call truncate(0), a small
simplification.

Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
13 years agorgw: remove some compilation warnings
Yehuda Sadeh [Thu, 25 Aug 2011 19:11:58 +0000 (12:11 -0700)]
rgw: remove some compilation warnings

13 years agorgw: don't override ret value in do_read_permisions
Yehuda Sadeh [Thu, 25 Aug 2011 19:06:16 +0000 (12:06 -0700)]
rgw: don't override ret value in do_read_permisions

13 years agorgw: return correct bucket missing error when needed
Yehuda Sadeh [Thu, 25 Aug 2011 18:25:19 +0000 (11:25 -0700)]
rgw: return correct bucket missing error when needed

13 years agorgw: fix 500 error on bucket creation
Yehuda Sadeh [Thu, 25 Aug 2011 18:11:17 +0000 (11:11 -0700)]
rgw: fix 500 error on bucket creation

13 years agorgw: fix clitest
Sage Weil [Thu, 25 Aug 2011 00:32:07 +0000 (17:32 -0700)]
rgw: fix clitest

13 years agoMonClient: fix possible null dereference
Colin Patrick McCabe [Thu, 25 Aug 2011 00:19:41 +0000 (17:19 -0700)]
MonClient: fix possible null dereference

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agomds/Server.c: fix memory leak in err path
Colin Patrick McCabe [Wed, 24 Aug 2011 23:36:12 +0000 (16:36 -0700)]
mds/Server.c: fix memory leak in err path

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agorgw: don't check for zero content when creating a bucket
Yehuda Sadeh [Thu, 25 Aug 2011 00:07:04 +0000 (17:07 -0700)]
rgw: don't check for zero content when creating a bucket

13 years agorgw: rgw_admin bucket stats dumps owner
Yehuda Sadeh [Thu, 25 Aug 2011 00:01:04 +0000 (17:01 -0700)]
rgw: rgw_admin bucket stats dumps owner

13 years agoclient: drop useless insert_trace timestamp argument
Sage Weil [Wed, 24 Aug 2011 23:54:02 +0000 (16:54 -0700)]
client: drop useless insert_trace timestamp argument

This is reachable via request.  It was also misnamed in the header :/.

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agoclient: be careful about replacing dentries during readdir assimilation
Sage Weil [Wed, 24 Aug 2011 23:51:15 +0000 (16:51 -0700)]
client: be careful about replacing dentries during readdir assimilation

When we are assimilate readdir results into our cache, we need to be more
careful about replacing existing dentries.  We were calling
insert_dentry_inode(), which would replace a name if it already exists,
which might include pd->first, an active iterator.

Move the dentry link/relink into the caller (where we already have an
iterator pointing to the existing item, if any).  Then update the dentry
lease information separately.

Fixes: #1391
Signed-off-by: Sage Weil <sage@newdream.net>
13 years agorgw: add radosgw_admin bucket stats
Yehuda Sadeh [Wed, 24 Aug 2011 23:41:10 +0000 (16:41 -0700)]
rgw: add radosgw_admin bucket stats

13 years agoclient: factor update_dentry_lease() out of insert_dentry_inode()
Sage Weil [Wed, 24 Aug 2011 23:25:49 +0000 (16:25 -0700)]
client: factor update_dentry_lease() out of insert_dentry_inode()

The first bit of insert_dentry_inode() handles the details of checking
whether an item still exists, un/relinking it, etc.

The second bit just updates the dentry lease information.

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agolibrados: fix memory leak in error path
Colin Patrick McCabe [Wed, 24 Aug 2011 22:51:38 +0000 (15:51 -0700)]
librados: fix memory leak in error path

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agotest/confutils.cc: fix resource leak
Colin Patrick McCabe [Wed, 24 Aug 2011 21:52:47 +0000 (14:52 -0700)]
test/confutils.cc: fix resource leak

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agoclient: carry inode reference across calls to _flush
Sage Weil [Wed, 24 Aug 2011 22:04:10 +0000 (15:04 -0700)]
client: carry inode reference across calls to _flush

Also drop the unused _flush Context* argument.

Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
13 years agorgw: fix prepare_get_obj when ENOENT
Yehuda Sadeh [Wed, 24 Aug 2011 22:02:01 +0000 (15:02 -0700)]
rgw: fix prepare_get_obj when ENOENT

13 years agorbd: fix shared_ptr usage
Colin Patrick McCabe [Wed, 24 Aug 2011 21:50:26 +0000 (14:50 -0700)]
rbd: fix shared_ptr usage

shared_ptr calls the disposal function even when the pointer is null
that is being disposed of.

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agodo_kernel_showmapped: fix resource leak
Colin Patrick McCabe [Wed, 24 Aug 2011 21:42:19 +0000 (14:42 -0700)]
do_kernel_showmapped: fix resource leak

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agoMerge remote branch 'origin/next'
Sage Weil [Wed, 24 Aug 2011 21:40:42 +0000 (14:40 -0700)]
Merge remote branch 'origin/next'

13 years agoclient: fix readdir + cap recovation/reissue race
Sage Weil [Wed, 24 Aug 2011 21:38:25 +0000 (14:38 -0700)]
client: fix readdir + cap recovation/reissue race

The current race:
 - we start readdir
 - mds revokes dir cap
 - some file gets removed
 - mds reissues dir cap
 - we finish readdir and set I_COMPLETE

We should only set it if there have been no FILE_SHARED reissues during
the readdir.

Note that we still set I_COMPLETE even if we don't have the cap; that's
useless but harmless, since it is undefined without FILE_SHARED being
set.

Signed-off-by: Sage Weil <sage@newdream.net>
13 years agorados_sync: fix resource leak on err
Colin Patrick McCabe [Wed, 24 Aug 2011 21:14:50 +0000 (14:14 -0700)]
rados_sync: fix resource leak on err

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agoCephBroker.cc: fix allocation/deallocation mismatch
Colin Patrick McCabe [Wed, 24 Aug 2011 21:14:25 +0000 (14:14 -0700)]
CephBroker.cc: fix allocation/deallocation mismatch

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agotest_rados_tool.sh: test export with --workers
Colin Patrick McCabe [Wed, 24 Aug 2011 17:51:04 +0000 (10:51 -0700)]
test_rados_tool.sh: test export with --workers

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agorados_sync: make number of worker threads configurable
Colin Patrick McCabe [Wed, 24 Aug 2011 17:41:44 +0000 (10:41 -0700)]
rados_sync: make number of worker threads configurable

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agorgw: clear warning
Yehuda Sadeh [Wed, 24 Aug 2011 20:17:20 +0000 (13:17 -0700)]
rgw: clear warning

13 years agoMerge branch 'wip-decouple-bucket'
Yehuda Sadeh [Wed, 24 Aug 2011 19:29:32 +0000 (12:29 -0700)]
Merge branch 'wip-decouple-bucket'

13 years agojournaler: fix waiting_for_zero clearing, again
Sage Weil [Wed, 24 Aug 2011 18:01:00 +0000 (11:01 -0700)]
journaler: fix waiting_for_zero clearing, again

We can only clear this when we have >= a period between flush_pos and
write_pos.

Clear the flag in _do_flush() so the check is not fragile, should this
ever be changed in the future.

Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
13 years agoRevert "pipe.c: #define GNU_SOURCE"
Sage Weil [Wed, 24 Aug 2011 16:38:09 +0000 (09:38 -0700)]
Revert "pipe.c: #define GNU_SOURCE"

This reverts commit ae9ba2249e372e7435e56f54fc2d8e22f3e42ff8.

This is handled in master by the Makefile.

13 years agoMerge branch 'next'
Sage Weil [Wed, 24 Aug 2011 16:38:01 +0000 (09:38 -0700)]
Merge branch 'next'

13 years agoclient: only link directories into hierarchy once
Sage Weil [Wed, 24 Aug 2011 16:31:22 +0000 (09:31 -0700)]
client: only link directories into hierarchy once

Directories can only link into the hierarchy once.  We assert as much
in readdir_r_cb().  Fix link() so that it unlinked the directory from the
old location when relinking somewhere new.  Be careful to do this after
we take inode refs to avoid any unpleasantness.

Fixes: #1429
Reported-by: Sam Lang <samlang@gmail.com>
Signed-off-by: Sage Weil <sage@newdream.net>
13 years agomon: fix store_state
Sage Weil [Wed, 24 Aug 2011 04:25:13 +0000 (21:25 -0700)]
mon: fix store_state

We need to bound the range we write to disk by m->last_committed; this got
lost in translation in commit dfab2c91f53289b1d527fb87126a639f5d5275f1.

Fixes paxos crashes in handle_begin

mon/Paxos.cc: In function 'void Paxos::handle_begin(MMonPaxos*)', in thread '0x7fc74d11f700'
mon/Paxos.cc: 393: FAILED assert(begin->last_committed == last_committed)

Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
13 years agojournaler: fix waiting_for_zero clearing
Sage Weil [Wed, 24 Aug 2011 00:33:44 +0000 (17:33 -0700)]
journaler: fix waiting_for_zero clearing

Clear this flag only if we know no subsequent flushes could be waiting on
a prezero operation.

Fixes MDS journaling hang under heavy journal load.

Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
13 years agoMerge branch 'master' into wip-decouple-bucket
Yehuda Sadeh [Wed, 24 Aug 2011 01:53:10 +0000 (18:53 -0700)]
Merge branch 'master' into wip-decouple-bucket

13 years agoadd missing include, required for certain environments
Yehuda Sadeh [Wed, 24 Aug 2011 01:52:35 +0000 (18:52 -0700)]
add missing include, required for certain environments

13 years agoMerge remote-tracking branch 'origin/master' into wip-decouple-bucket
Yehuda Sadeh [Wed, 24 Aug 2011 01:52:08 +0000 (18:52 -0700)]
Merge remote-tracking branch 'origin/master' into wip-decouple-bucket

Conflicts:
src/rgw/rgw_rados.cc

13 years agorgw: fix multipart upload
Yehuda Sadeh [Wed, 24 Aug 2011 01:33:12 +0000 (18:33 -0700)]
rgw: fix multipart upload

don't try to reset meta object with category

13 years agoparallelize rados export and import
Colin Patrick McCabe [Tue, 23 Aug 2011 17:45:29 +0000 (10:45 -0700)]
parallelize rados export and import

* use workqueues to parallelize rados export and import

* Put export and import into separate files

* Fix RADOS_SYNC_TMP_SUFFIX_LEN bug

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agoWorkQueue: don't be chatty about finishing workers
Colin Patrick McCabe [Tue, 23 Aug 2011 18:13:03 +0000 (11:13 -0700)]
WorkQueue: don't be chatty about finishing workers

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agotest_rados_tool.sh: fix xattr test
Colin Patrick McCabe [Tue, 23 Aug 2011 18:12:38 +0000 (11:12 -0700)]
test_rados_tool.sh: fix xattr test

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agolibrados: fix IoCtx copy constructor
Colin Patrick McCabe [Tue, 23 Aug 2011 17:57:04 +0000 (10:57 -0700)]
librados: fix IoCtx copy constructor

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agorados_sync:don't need to declare CHANGED_XATTRS 2x
Colin Patrick McCabe [Mon, 22 Aug 2011 19:55:55 +0000 (12:55 -0700)]
rados_sync:don't need to declare CHANGED_XATTRS 2x

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agoSome fixes to double-dash support, fix warning
Colin Patrick McCabe [Wed, 24 Aug 2011 01:12:06 +0000 (18:12 -0700)]
Some fixes to double-dash support, fix warning

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agorgw: if pool is empty recreate
Yehuda Sadeh [Wed, 24 Aug 2011 01:06:02 +0000 (18:06 -0700)]
rgw: if pool is empty recreate

13 years agoceph_argparse: handle double dashes consistently
Colin Patrick McCabe [Tue, 23 Aug 2011 23:30:03 +0000 (16:30 -0700)]
ceph_argparse: handle double dashes consistently

Handle double dashes in the ceph_argparse functions, so that any piece
of code doing argument parsing will correctly interpret them.

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agoinjectargs: binary flags only consume 1 token
Colin Patrick McCabe [Tue, 23 Aug 2011 22:43:52 +0000 (15:43 -0700)]
injectargs: binary flags only consume 1 token

Never consume more than one token when parsing a binary flag.

Basically, your choices are --foo=false, --foo=true, and just --foo.
However, only 1 token will ever be examined.

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agoconfig: print warning on injectargs parse fail
Colin Patrick McCabe [Tue, 23 Aug 2011 21:46:05 +0000 (14:46 -0700)]
config: print warning on injectargs parse fail

injectargs should print a warning message when its parser fails.

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agoinjectargs test: test foo=true
Colin Patrick McCabe [Tue, 23 Aug 2011 21:34:08 +0000 (14:34 -0700)]
injectargs test: test foo=true

Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
13 years agomon: allow wildcards for mon cap command matching
Sage Weil [Tue, 23 Aug 2011 22:18:29 +0000 (15:18 -0700)]
mon: allow wildcards for mon cap command matching

We need wildcards, actually.  * is any token, ... is 1 or more tokens.

Signed-off-by: Sage Weil <sage@newdream.net>