]>
git.apps.os.sepia.ceph.com Git - ceph.git/log
Colin Patrick McCabe [Wed, 25 May 2011 21:22:51 +0000 (14:22 -0700)]
pybind/rados.py: rados.Object.key should be string
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Wed, 25 May 2011 19:58:39 +0000 (12:58 -0700)]
obysnc: RgwStore: make sure destination users exist
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Wed, 25 May 2011 19:36:37 +0000 (12:36 -0700)]
obsync: fix DST_OWNER
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Wed, 25 May 2011 19:05:15 +0000 (12:05 -0700)]
obsync: Add boto_retries, remove rgw_store.prefix
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Wed, 25 May 2011 18:01:16 +0000 (11:01 -0700)]
rados python bindings: handle xattrs with NULL
Handle extended attributes that contain NULL bytes correctly, rather
than treating everything as zero-terminated C strings.
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Wed, 25 May 2011 17:50:15 +0000 (10:50 -0700)]
RgwStore: fix some ACL issues
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Tue, 24 May 2011 22:39:58 +0000 (15:39 -0700)]
test-obysnc.py: support librgw testing
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Tue, 24 May 2011 21:15:53 +0000 (14:15 -0700)]
Rename RadosStore to RgwStore
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Thu, 19 May 2011 23:33:13 +0000 (16:33 -0700)]
test-obsync: refactor a little bit
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Wed, 18 May 2011 23:39:18 +0000 (16:39 -0700)]
Proper ACL support for rados targets
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Sage Weil [Wed, 25 May 2011 04:14:59 +0000 (21:14 -0700)]
crushtool: clean up add-item a bit; don't add item to same bucket twice
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Wed, 25 May 2011 04:05:47 +0000 (21:05 -0700)]
crushtool: fix remove-item
Scan all buckets instead of doing a tree traverse.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Wed, 25 May 2011 03:30:38 +0000 (20:30 -0700)]
radosgw_admin: update clitest
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Colin Patrick McCabe [Wed, 25 May 2011 01:16:08 +0000 (18:16 -0700)]
mkcephfs.in: print out usage if no actions given
If the user didn't specify any actions, print out a usage message rather
than silently exiting.
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Wed, 25 May 2011 00:50:24 +0000 (17:50 -0700)]
rgw: Fix RGWAccess::init_storage_provider
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Sage Weil [Wed, 25 May 2011 00:05:30 +0000 (17:05 -0700)]
mkcephfs: error out on bad usage
Signed-off-by: Sage Weil <sage@newdream.net>
Yehuda Sadeh [Tue, 24 May 2011 23:40:12 +0000 (16:40 -0700)]
make: fix build for rgw
Yehuda Sadeh [Tue, 24 May 2011 23:33:11 +0000 (16:33 -0700)]
rgw_admin: clean warning
Yehuda Sadeh [Tue, 24 May 2011 22:30:17 +0000 (15:30 -0700)]
Merge commit 'origin/master' into rgw-multiuser
Yehuda Sadeh [Tue, 24 May 2011 21:29:50 +0000 (14:29 -0700)]
rgw_admin: add key create
Yehuda Sadeh [Tue, 24 May 2011 21:17:59 +0000 (14:17 -0700)]
rgw_admin: subuser and key removal
Colin Patrick McCabe [Tue, 24 May 2011 19:36:07 +0000 (12:36 -0700)]
test_common.sh: skip rm before put
The rm before the put is unecessary and actually incorrect now.
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Tue, 24 May 2011 19:34:56 +0000 (12:34 -0700)]
radostool: rados put should use write_full
If "rados put" uses write instead of write_full, the resulting object on
the server may be a mismash of old and new objects, if the old object
was longer than the new one. This is fairly counterintuitive behavior
for radostool, so remove it.
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Tue, 24 May 2011 19:22:30 +0000 (12:22 -0700)]
Merge branch 'wip_ceph_context'
Colin Patrick McCabe [Mon, 23 May 2011 23:25:57 +0000 (16:25 -0700)]
Create a libcommon service thread
Create a libcommon service thread. Use it to handle SIGHUP.
Handle it by means of a flag that gets set. Using a queue would raise
the complicated question of what to do when the queue was full.
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Sage Weil [Tue, 24 May 2011 17:00:23 +0000 (10:00 -0700)]
librados: len should be size_t
Unsigned, and size_t because it's a buffer size.
Fixes signedness warning in testrados.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Tue, 24 May 2011 16:26:40 +0000 (09:26 -0700)]
mds: clean up get_or_create_stray
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Tue, 24 May 2011 16:17:24 +0000 (09:17 -0700)]
Merge branch 'stable'
Sage Weil [Tue, 24 May 2011 04:11:44 +0000 (21:11 -0700)]
v0.28.1
Colin Patrick McCabe [Mon, 23 May 2011 21:02:15 +0000 (14:02 -0700)]
librads, libceph: store CephContext
Don't use the global g_ceph_context. Instead, store the CephContext in
the structures provided by the library user.
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Mon, 23 May 2011 17:11:15 +0000 (10:11 -0700)]
Add CephContext
A CephContext represents the context held by a single library user.
There can be multiple CephContexts in the same process.
For daemons and utility programs, there will be only one CephContext.
The CephContext contains the configuration, the dout object, and
anything else that you might want to pass to libcommon with every
function call.
Move some non-config things out of md_config_t and into CephContext.
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Mon, 23 May 2011 23:29:49 +0000 (16:29 -0700)]
Split common_init_daemonize from common_init_finish
Split off common_init_daemonize from common_init_finish. cfuse is a
daemon that calls common_init_finish, but handles daemonization itself.
This fixes cfuse.
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Yehuda Sadeh [Mon, 23 May 2011 23:52:59 +0000 (16:52 -0700)]
rgw_admin: make interface a bit more explicit
Yehuda Sadeh [Mon, 23 May 2011 22:12:48 +0000 (15:12 -0700)]
rgw: subuser permissions
Sage Weil [Mon, 23 May 2011 21:58:26 +0000 (14:58 -0700)]
mon: verify that crush max does not exceed osd max
- when injecting a new crushmap
- when adjusting osdmap max_osd
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sun, 22 May 2011 23:25:35 +0000 (16:25 -0700)]
crushtool: add --reweight-item <name> <weight>
Reweight and individual item via crushtool.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Sat, 21 May 2011 19:55:16 +0000 (12:55 -0700)]
osdmaptool: fail --import-crush if crush max_devices > osdmap max_osd
Crush will spew non-deterministic badness if it walks off the end of
the osd_weight vector.
Signed-off-by: Sage Weil <sage@newdream.net>
Colin Patrick McCabe [Fri, 20 May 2011 23:35:52 +0000 (16:35 -0700)]
common_init: don't init crypto until after fork
Get rid of the initialize-then-shutdown-crypto hack. We just initialize
crypto once, after it is safe to do so. There is now a single callback,
common_init_finish, which does the final stage of initialization,
including starting crypto and daemonization (if required.)
common_init_finish needs to be done before messenger::start().
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Fri, 20 May 2011 22:12:49 +0000 (15:12 -0700)]
ceph_crypto: add assert_init
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Sage Weil [Sat, 21 May 2011 01:16:49 +0000 (18:16 -0700)]
config: delete after new
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Sat, 21 May 2011 00:10:15 +0000 (17:10 -0700)]
crush: fix signedness warnings
Signed-off-by: Sage Weil <sage@newdream.net>
Yehuda Sadeh [Fri, 20 May 2011 23:46:14 +0000 (16:46 -0700)]
rgw_admin: able to create multiple keys/subusers
Sage Weil [Fri, 20 May 2011 23:45:57 +0000 (16:45 -0700)]
crushtool: --remove-item name
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 20 May 2011 23:41:16 +0000 (16:41 -0700)]
crush: fix tree bucket encoding
I wonder how long this has been broken!
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 20 May 2011 23:40:36 +0000 (16:40 -0700)]
crush: fix tree weight accessor, decompile
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 20 May 2011 22:44:15 +0000 (15:44 -0700)]
crushtool: default to hash 0 (rjenkins1)
Otherwise we get 255 which is undefined and get bad results!
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Yehuda Sadeh [Fri, 20 May 2011 22:15:48 +0000 (15:15 -0700)]
rgw: user info structure supports multiple subusers and keys
Sage Weil [Fri, 20 May 2011 22:08:06 +0000 (15:08 -0700)]
osd: update last_epoch_clean in PG::Info::History::merge()
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 22:04:57 +0000 (15:04 -0700)]
osd: small cleanup
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 22:04:09 +0000 (15:04 -0700)]
osd: merge history when primary sends replica new pg info
This, among other things, lets us update last_epoch_started and
last_epoch_clean.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 21:45:36 +0000 (14:45 -0700)]
osd: more heartbeat rework
A few things:
- track Connection* instead of entity_inst_t for hb peers
- we can only send maps over the cluster_messenger
- if peer is still alive, do that
- if peer is not, send dying MOSDPing ping with YOU_DIED flag
Sage Weil [Fri, 20 May 2011 21:43:57 +0000 (14:43 -0700)]
msgr: don't close close_on_empty until outgoing messages are acked
Otherwise, if we close the socket, we may lose in-flight data.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 20:25:22 +0000 (13:25 -0700)]
osd: only forget peer epochs if they are down AND no longer heartbeat peers
If we forget the peer epoch when we see them go down, we won't share the
map later in update_heartbeat_peers() to tell them they're down.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 20:01:22 +0000 (13:01 -0700)]
osd: show last_epoch_clean in PG::Info::History printer
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 19:55:29 +0000 (12:55 -0700)]
osd: rework peer map epoch caching
We try to keep track of which epochs our peers have so that we can be
semi-intelligent about which map incrementals we send preceeding any
messages. Since this is useful from the heartbeat and cluster channels/
threads, protect the data with an inner lock and clean up the callers.
Be smarter about when we forget.
Make note of peer epoch when we receive a ping.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 19:22:22 +0000 (12:22 -0700)]
mon: fix parsing of 'osd foo N ...' commands with multiple ids
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Colin Patrick McCabe [Fri, 20 May 2011 21:23:10 +0000 (14:23 -0700)]
dout: reopen log files on SIGHUP
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Fri, 20 May 2011 21:23:10 +0000 (14:23 -0700)]
dout: reopen log files on SIGHUP
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Colin Patrick McCabe [Fri, 20 May 2011 18:35:19 +0000 (11:35 -0700)]
Add SignalSafeQueue
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Sage Weil [Fri, 20 May 2011 18:29:05 +0000 (11:29 -0700)]
osd: clean up old _from target cleanup; fix one case; share map
Clean up the code to mirror the _to case.
Previously we would not mark down an old _from that is still a _to but with
a new address. Now we do.
Share a map while we're at it, just to be nice!
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 18:25:27 +0000 (11:25 -0700)]
osd: mark down old _to targets
If a peer remains a _to target but their address changes, we still want
to mark down the old connection.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 18:20:20 +0000 (11:20 -0700)]
osd: share map with old _to peers
Use new msgr hooks to do this cleanly.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 18:17:19 +0000 (11:17 -0700)]
osd: clean up handle_osd_ping output
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 17:54:46 +0000 (10:54 -0700)]
osd: ignore stale requests for heartbeats
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 17:43:12 +0000 (10:43 -0700)]
osd: don't prioritize heartbeat requests
This could conceivably screw up ordering, and priority doesn't matter
anyway when this is the first message we send to this peer.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 17:42:16 +0000 (10:42 -0700)]
osd: do not clobber explicitly requested heartbeat_to target addresss
Consider peer P.
- P does down in, say, epoch 60, and back up in epoch 70
- P and requests a heartbeat, as_of 70
- We update to map 50, and coincidentally add the same peer as a target
- We set the heartbeat_to[P] = 50 and start sending to the _old_ address
- P marks us down because we stop sending to the new addr
- We eventually get map 70, but it's too late!
Make sure we preserve any _to targets _and_ their epoch+inst.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 16:29:10 +0000 (09:29 -0700)]
osd: request proper log extent for missing
We can't blinding ask for everything since last_epoch_started because that
may mean we get some fragment of a backlog. Look at the peer's log
ranges and request the correct thing. Also, in fulfill_log, infer what
the primary should have asked for if they make a bad request.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 15:44:42 +0000 (08:44 -0700)]
osd: fix log bounds check
We weren't accounting for the case where we have
(foo,foo]+backlog
i.e., everything is backlog, and rbegin().version != log.head.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 15:35:44 +0000 (08:35 -0700)]
osd: osd# is in log entry header/prefix
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 15:33:07 +0000 (08:33 -0700)]
osd: log broken pg state to monitor on startup, activate
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 15:09:11 +0000 (08:09 -0700)]
osd: fix proc_replica_log when peer log is empty
If the peer log is empty, and we break out of the loop on the first pass,
then clearly last_update has not been adjusted.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 14:25:24 +0000 (07:25 -0700)]
osd: encode keyring as plaintext after --mkkey
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 14:25:16 +0000 (07:25 -0700)]
keyring: make encode_plaintext method
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 07:41:31 +0000 (00:41 -0700)]
Merge branch 'wip_choose_acting' into stable
Sage Weil [Fri, 20 May 2011 07:27:00 +0000 (00:27 -0700)]
osd: take remote log when it is clearly superior
I'm hitting a case where the primary is compensating for a replica's
last_complete < log.tail by sending a log+backlog, but the replica
isn't smart enough to take advantage. In this case,
replica: log(781'26629,781'26631]
from primary: log(781'26629,781'26631]+backlog
result: log(781'26629,781'26631]
Doh!
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 07:14:24 +0000 (00:14 -0700)]
osd: fix compensation for bad last_complete
If the peer has a last_complete below their tail, we can get by with our
log (without backlog) if our tail if _before_ their last_complete, not
after. Otherwise, we need a backlog!
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 06:48:53 +0000 (23:48 -0700)]
osd: remove some build_prior stringstream cruft
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 06:46:19 +0000 (23:46 -0700)]
osd: remove useless debug print
We dump this (and more) at the end of the PgPriorSet constructor.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 06:40:12 +0000 (23:40 -0700)]
osd: include past acting osds if they were up
This fixes a bug where we were excluding up (but not acting) nodes from
past intervals, which in turn was triggering a nasty choose_acting loop
(because we _do_ already include acting but !up from the current
interval).
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 06:38:25 +0000 (23:38 -0700)]
osd: do not exclude me during build_prior
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 06:25:32 +0000 (23:25 -0700)]
osd: show final build_prior result
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 20 May 2011 03:45:48 +0000 (20:45 -0700)]
mon: log mkfs as INFO with fs
The [ERR] log level is misleading.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Josh Durgin [Fri, 20 May 2011 00:19:59 +0000 (17:19 -0700)]
OSD, PG: ignore peering messages from before the last peering restart
Check them before entering the state machine so we can
safely enter the Crashed state on unexpected messages
from the current interval.
Signed-off-by: Josh Durgin <josh.durgin@dreamhost.com>
Josh Durgin [Fri, 20 May 2011 00:46:40 +0000 (17:46 -0700)]
OSD: decrement message refcount before returning
Signed-off-by: Josh Durgin <josh.durgin@dreamhost.com>
Sage Weil [Fri, 20 May 2011 00:20:18 +0000 (17:20 -0700)]
mds: kick linklock on revoke_stale_caps
Also use the eval() method and issue caps instead of calling the individual
eval methods.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Thu, 19 May 2011 23:15:59 +0000 (16:15 -0700)]
debian: no shlibs:Depends for obsync either
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Thu, 19 May 2011 23:15:26 +0000 (16:15 -0700)]
debian: no shlibs:Depends for -dev packages
So says dpkg-gencontrol, at least:
warning: dpkg-gencontrol: Depends field of package librados-dev: unknown substitution variable ${shlibs:Depends}
...
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Thu, 19 May 2011 23:13:34 +0000 (16:13 -0700)]
librbd: don't need to link against crypto libs
All that is done by librados.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Samuel Just [Thu, 19 May 2011 21:13:56 +0000 (14:13 -0700)]
PG: add_event, add_next_event: ignore prior_version on backlog events
We would not have the previous version if we are merging backlog events.
Signed-off-by: Samuel Just <samuel.just@dreamhost.com>
Brian Chrisman [Thu, 19 May 2011 20:22:33 +0000 (13:22 -0700)]
expanding testceph to test open/readdir/telldir
Signed-off-by: Brian Chrisman <brchrisman@gmail.com>
Signed-off-by: Sage Weil <sage@newdream.net>
Brian Chrisman [Thu, 19 May 2011 20:22:32 +0000 (13:22 -0700)]
add ceph_readdir() to libceph
Signed-off-by: Brian Chrisman <brchrisman@gmail.com>
Signed-off-by: Sage Weil <sage@newdream.net>
Colin Patrick McCabe [Thu, 19 May 2011 21:27:19 +0000 (14:27 -0700)]
librados: add python bindings for getxattrs
Add python bindings for getxattrs. Test getxattr, getxattrs, and
setxattr.
Signed-off-by: Colin McCabe <colin.mccabe@dreamhost.com>
Sage Weil [Thu, 19 May 2011 22:03:13 +0000 (15:03 -0700)]
client: hold FILE_BUFFER ref while waiting for dirty throttle
We may block in the write path because we've reached out dirty data limit.
Hold a reference to the FILE_BUFFER cap during that interval so we don't
lose the cap and put new dirty buffers into the objectcacher out of turn.
(We could also recheck our ability to take the ref after blocking, but I
think this is cleaner.)
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Thu, 19 May 2011 22:01:50 +0000 (15:01 -0700)]
client: clean up _flush callers
Have _flush return true if there are no dirty buffers. Clean up some
redundant conditionals in the callers
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Thu, 19 May 2011 22:00:34 +0000 (15:00 -0700)]
client: assert(in) on _flush
We should never arrive in _flush() and not have a reference to the inode
in question, because the presence of dirty buffers pins the inode. This
condition was introduced forever ago; clean it out.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Thu, 19 May 2011 21:50:41 +0000 (14:50 -0700)]
client: be more careful with FILE_BUFFER cap refs
We should either hold a ref or not; whether we release one can't depend on
whether one is held because we can't assume the ref belongs to us.
This changes the fix in
cf6b1de4 so that the ObjectCacher just calls the
flush callback if it happens to trim all dirty buffers.
We also drop the (bogus) assert about the number of refs held.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Thu, 19 May 2011 19:21:51 +0000 (12:21 -0700)]
client: _flush should no-op if nothing to flush
If there are no FILE_BUFFER cap_refs, then we can bail out early.
Otherwise we will end up dropping refs we don't have.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Thu, 19 May 2011 22:04:19 +0000 (15:04 -0700)]
Merge remote branch 'origin/stable'
Josh Durgin [Thu, 19 May 2011 21:31:30 +0000 (14:31 -0700)]
PG: choose_log_location: prefer OSDs with a backlog
Without preferring an OSD with a backlog, PGs would get stuck in the
active state when acting != up and the backlog was on an OSD with the
same last_update but a lower number or log_tail.
Signed-off-by: Josh Durgin <josh.durgin@dreamhost.com>
Josh Durgin [Thu, 19 May 2011 17:44:57 +0000 (10:44 -0700)]
OSD: send a log in response to a log query when the pg dne
Signed-off-by: Josh Durgin <josh.durgin@dreamhost.com>