]>
git.apps.os.sepia.ceph.com Git - ceph.git/log
Sage Weil [Tue, 22 Nov 2011 22:53:45 +0000 (14:53 -0800)]
mon: name self in monmap if --public-addr specified during mkfs
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Tue, 22 Nov 2011 18:09:41 +0000 (10:09 -0800)]
mon: mark down all connections when rank changes
The election and some other stuff depend on msg->get_source().num() to get
the peer rank, and that is part of the connection state. If it changes,
we need to close old connections and open new ones so that we aren't
taken for someone else (like mon.-1).
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Tue, 22 Nov 2011 18:08:48 +0000 (10:08 -0800)]
mon: handle rank change in bootstrap
The rank can change either because we probe and get a new monmap, or
because we get one via paxos. Move the checks to bootstrap() to catch
both cases.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Tue, 22 Nov 2011 17:53:52 +0000 (09:53 -0800)]
mon: pick an address when joining and existing cluster
If we are joining an existing cluster, we can pick whatever address we
want (e.g., one specified by public_addr or public_network).
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Tue, 22 Nov 2011 17:52:58 +0000 (09:52 -0800)]
mon: remove unused myaddr
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Tue, 22 Nov 2011 17:52:52 +0000 (09:52 -0800)]
mon: simplify suicide when removed from map
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Tue, 22 Nov 2011 00:17:21 +0000 (16:17 -0800)]
Merge remote branch 'gh/subnet'
Sage Weil [Tue, 22 Nov 2011 00:00:34 +0000 (16:00 -0800)]
Merge remote branch 'gh/wip-mon'
Tommi Virtanen [Mon, 21 Nov 2011 21:32:45 +0000 (13:32 -0800)]
mds, osd, synclient: Pick cluster_addr/public_addr based on *_network.
Instead of specifying an IP address in ceph.conf like
[global]
cluster_addr = 10.1.2.3
you can now avoid the node-specific configuration and just say
[global]
cluster_network = 10.1.2.0/24
The *_network variables can also take a whitespace-separated list of
networks, to be checked in that order:
[global]
cluster_network = 10.1.2.0/24 192.168.42.192/26
Tommi Virtanen [Sat, 19 Nov 2011 00:55:29 +0000 (16:55 -0800)]
common/pickaddr: Pick cluster_addr/public_addr based on *_network.
Tommi Virtanen [Sat, 19 Nov 2011 00:47:45 +0000 (16:47 -0800)]
common/ipaddr: Add utility function to parse ip/cidr style networks.
Signed-off-by: Tommi Virtanen <tommi.virtanen@dreamhost.com>
Tommi Virtanen [Wed, 16 Nov 2011 21:39:29 +0000 (13:39 -0800)]
common/ipaddr: Find a configured IP address in given subnet.
Signed-off-by: Tommi Virtanen <tommi.virtanen@dreamhost.com>
Tommi Virtanen [Mon, 21 Nov 2011 18:12:29 +0000 (10:12 -0800)]
msg: Move public_addr use outside ->bind()
Tommi Virtanen [Wed, 16 Nov 2011 21:40:02 +0000 (13:40 -0800)]
common/str_list: Make unused return value void.
Signed-off-by: Tommi Virtanen <tommi.virtanen@dreamhost.com>
Tommi Virtanen [Sat, 19 Nov 2011 00:55:34 +0000 (16:55 -0800)]
osd: Remove unused variable.
Sage Weil [Mon, 21 Nov 2011 21:28:36 +0000 (13:28 -0800)]
osd: fix 'stop' command
Special case. We can't join the command_tp thread from itself.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Mon, 21 Nov 2011 21:23:59 +0000 (13:23 -0800)]
osd: protect handle_osd_map requeueing with queue lock
pending_ops was protected by osd_lock, but it tracks something in the
queue, which has it's own lock. Messy. Also, useless, since
wait_for_no_ops had a single caller in shutdown() that op_wq.drain() can
do for us.
Rip it out, and track queue size under the queue lock.
Fixes: #1727
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Mon, 21 Nov 2011 19:15:38 +0000 (11:15 -0800)]
osd: lock pg when requeuing requests
The op queue is shut down, so this is mostly safe, unless someone comes
through and does requeue_ops() from a callback or something.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Mon, 21 Nov 2011 18:33:53 +0000 (10:33 -0800)]
paxosservice: tolerate _active() call when not active
This can happen when multiple C_Active events are queued, and the first
does a propose_pending() (moving us into updating state).
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Thu, 17 Nov 2011 20:08:40 +0000 (12:08 -0800)]
objecter: simplify map request check
We should request a missing/intervening map if it appears to exist.
Otherwise, skip it.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Mon, 21 Nov 2011 17:19:26 +0000 (09:19 -0800)]
objecter: cancel tick event on shutdown
Hopefully this is the root cause for
2011-11-20 23:57:41.555292
7f75dd743780 ceph version
0.38-205-g3b53b72
(commit:
3b53b722b34b5284e6b8a5571a08d4b7ec276241 ), process ceph-fuse, pid
21223
* Caught signal (Segmentation fault) *
in thread
7f75d9c6e700
ceph version
0.38-205-g3b53b72
(commit:
3b53b722b34b5284e6b8a5571a08d4b7ec276241 )
1: /tmp/cephtest/binary/usr/local/bin/ceph-fuse() [0x6993a4]
2: (()+0xfb40) [0x7f75dd0eeb40]
3: (PerfCounters::set(int, unsigned long)+0x2a) [0x511bca]
4: (Objecter::tick()+0x1f3) [0x653f43]
5: (Objecter::C_Tick::finish(int)+0x15) [0x66aef5]
6: (SafeTimer::timer_thread()+0x4b0) [0x5825c0]
7: (SafeTimerThread::entry()+0x15) [0x586865]
8: (Thread::_entry_func(void)+0x12) [0x52a832]
9: (()+0x7971) [0x7f75dd0e6971]
10: (clone()+0x6d) [0x7f75db97592d]
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sun, 20 Nov 2011 22:26:09 +0000 (14:26 -0800)]
paxos: fix sharing of learned commits during collect/last
We can learn either an uncommitted or committed value during the
collect/last recovery phase. For the committed values, we need to remember
each peer's first/last_committed and share only at the end to avoid a
situation like:
- mon.1 has same last_committed as us
- mon.2 has newer last_commited, we save it
- mon.3 has same last_commited as mon.1, we share new value
- done... but mon.1 never got mon.2's newer commit.
Instead, save the commit sharing until the collect process completes, so
we know that any committed value learned from anyone is shared with
everyone who needs it.
This fixes a crash like
mon/Paxos.cc: In function 'void Paxos::handle_begin(MMonPaxos*)', in thread '
7fd91192c700 '
mon/Paxos.cc: 400: FAILED assert(begin->last_committed == last_committed)
ceph version
0.38-208-g9aabd39 (commit:
9aabd3982cceb7e8489412b4bfbb4c2387880de2 )
1: (ceph::__ceph_assert_fail(char const*, char const*, int, char const*)+0x76) [0x72454e]
2: (Paxos::handle_begin(MMonPaxos*)+0x363) [0x6499ef]
3: (Paxos::dispatch(PaxosServiceMessage*)+0x2b4) [0x64db2c]
4: (Monitor::_ms_dispatch(Message*)+0xdc6) [0x6205c2]
5: (Monitor::ms_dispatch(Message*)+0x3a) [0x62831a]
6: (Messenger::ms_deliver_dispatch(Message*)+0x63) [0x7d1f31]
7: (SimpleMessenger::dispatch_entry()+0x7c2) [0x7bb786]
8: (SimpleMessenger::DispatchThread::entry()+0x2c) [0x6070fa]
9: (Thread::_entry_func(void*)+0x23) [0x6f3f69]
10: (()+0x7971) [0x7fd9153a1971]
11: (clone()+0x6d) [0x7fd913c3092d]
Signed-off-by: Sage Weil <sage@newdream.net>
Yehuda Sadeh [Sun, 20 Nov 2011 21:17:04 +0000 (13:17 -0800)]
rgw: support alternative date formatting
being used by s3cmd
Sage Weil [Fri, 18 Nov 2011 18:35:44 +0000 (10:35 -0800)]
paxosservice: consolidate _active and _commit
Use the same callback for when paxos goes active and for when it commits
something. The response in both cases is the same.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 18 Nov 2011 18:05:35 +0000 (10:05 -0800)]
paxosservice: remove unused committed() callback
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 18 Nov 2011 18:01:30 +0000 (10:01 -0800)]
mon: mdsmon: tick() from on_active() instead of committed()
Same effect, and avoids useless committed().
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 18 Nov 2011 17:56:10 +0000 (09:56 -0800)]
mon: share random osd map from update_from_paxos, not committed()
This will let us remove committed() entirely.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 18 Nov 2011 19:04:24 +0000 (11:04 -0800)]
config: support --no-<foo> for bool options
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 18 Nov 2011 19:04:09 +0000 (11:04 -0800)]
config: whitespace
Signed-off-by: Sage Weil <sage@newdream.net>
Greg Farnum [Fri, 18 Nov 2011 23:56:35 +0000 (15:56 -0800)]
osdmon: set the maps-to-keep floor to be at least epoch 0
Looks like this conditional was just set backwards by mistake. There
have been a number of issues with OSDMap versions that are probably
related to this...
(Thanks to some smarts in trim_to, we at least did not trim ALL of
our maps. But on every tick prior to epoch 500 [that's the default]
the leader was trimming all old maps off the system.)
Signed-off-by: Greg Farnum <gregory.farnum@dreamhost.com>
Samuel Just [Fri, 18 Nov 2011 01:17:06 +0000 (17:17 -0800)]
Revert "osd: simplify finalizing scrub on replica"
This reverts commit
dd5087fabb2a743741a96ee4610379afa8431f68 .
Calling osr.flush() is not quite enough since the onreadable callbacks
may not have been called (thus, last_update_applied may still lag behind
the tail of the log).
Signed-off-by: Samuel Just <samuel.just@dreamhost.com>
Samuel Just [Thu, 17 Nov 2011 21:45:08 +0000 (13:45 -0800)]
FileStore.cc: onreadable callbacks in OpSequencer order is enough
Signed-off-by: Samuel Just <samuel.just@dreamhost.com>
Greg Farnum [Fri, 18 Nov 2011 16:49:35 +0000 (08:49 -0800)]
osd: error responses should trigger all requested notifications.
There's no good reason I can find to limit error code responses to
the ACK.
Signed-off-by: Greg Farnum <gregory.farnum@dreamhost.com>
Greg Farnum [Fri, 18 Nov 2011 16:47:09 +0000 (08:47 -0800)]
objecter: trigger oncommit acks if the request returns an error code.
Many users only set oncommit acks, so if they get an error code
(which comes only as a CEPH_OSD_OP_ACK right now) the request
disappears into the ether.
(And remove stupid debug statements while we're at it.)
Signed-off-by: Greg Farnum <gregory.farnum@dreamhost.com>
Sage Weil [Fri, 18 Nov 2011 17:49:03 +0000 (09:49 -0800)]
paxos: do not create_pending if !active
This avoids a scenario like:
- _active()
- proposes value
- _commit()
- creates new pending, even though in updating state
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 18 Nov 2011 17:43:09 +0000 (09:43 -0800)]
Revert "mon: don't propose new state from update_from_paxos"
This reverts commit
66c628acc8be71a92e801179431e4b938b857b3d .
Sage Weil [Fri, 18 Nov 2011 04:45:54 +0000 (20:45 -0800)]
mon: don't propose new state from update_from_paxos
Proposing a new state from within update_from_paxos() confuses some callers,
like PaxosService::_active(). Instead, do it in the on_active() callback.
This also let's us collapse the check_osd_map() caller into on_active(),
and makes it happen on leaders and peons alike, which ought to avoid some
of the pg creation lag we see sometimes (presumably when the osds have
sessions with peons instead of the leader).
Fixes: #1708
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Yehuda Sadeh [Fri, 18 Nov 2011 00:54:51 +0000 (16:54 -0800)]
rgw: if swift url is not set up, just use whatever client used
Sage Weil [Thu, 17 Nov 2011 23:01:17 +0000 (15:01 -0800)]
fuse: fix readdir return code
Ignore ENOSPC generated by our own callback, as it is only used to
terminate the loop.
Broken by commit
cd90061239a598f6fca94326b6d2c32f325c96eb .
Fixes: #1728
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Thu, 17 Nov 2011 22:11:38 +0000 (14:11 -0800)]
paxos: fix trimming when we skip over incrementals
Remove open-coded trimming of old states and use our method (that also
removes additional per-state files). Fixes old stray state files.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Thu, 17 Nov 2011 22:10:34 +0000 (14:10 -0800)]
paxos: store stashed state _and_ incrementals
Paxos::share_state() may share a stashed state and incrementals that
follow; we need to store the same.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 05:58:53 +0000 (21:58 -0800)]
mon: elector: always start election via monitor
Don't go from active -> electing without passing (monitor) go.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Thu, 17 Nov 2011 20:07:34 +0000 (12:07 -0800)]
common: libraries should not log to stdout/stderr
Certainly not by default.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Thu, 17 Nov 2011 19:56:37 +0000 (11:56 -0800)]
objecter: set skipped_map if we skip a map
This ensures that we resend _all_ requests, since we aren't sure which
may have mapped to a different primary and then back. This was missed in
the original implementation in
4fe9cca5dd63a1924be2b5cb18f542fb4b97a768 .
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Thu, 17 Nov 2011 19:39:55 +0000 (11:39 -0800)]
objecter: add is_locked() asserts
Sanity check.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Thu, 17 Nov 2011 19:39:36 +0000 (11:39 -0800)]
objecter: send slow osd MPing via Connection*
This may address #1732 indirectly because we have a Connection* reference
here. However, it's still not clear how we ended up with an OSDSession*
for an osd that doesn't exist. :/
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Wed, 16 Nov 2011 21:10:58 +0000 (13:10 -0800)]
osd: add pending_ops assert
Just a sanity check, hopefully helping us track down #1727.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Wed, 16 Nov 2011 19:01:59 +0000 (11:01 -0800)]
mon: renamed get_latest* -> get_stashed*
This makes e.g. get_latest_version() vs get_last_committed() less
confusing.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Wed, 16 Nov 2011 18:57:23 +0000 (10:57 -0800)]
mon: fix ver tracking for auth database
Local variable keys_ver needs to be updated when we slurp up latest stashed
version.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Wed, 16 Nov 2011 18:54:59 +0000 (10:54 -0800)]
mon: always load stashed version when version doesn't match
The slurp process can happen after the monitor has started and has some
in-memory version of the state, and that process may wipe out old
incrementals and change the stashed version. That means that in
update_from_paxos, we need to pull the stashed version if it doesn't
match what we currently have or else we may not have the incrementals we
need to get up to date.
This simplifies and cleans up that code a bit so it is not specific to
monitor startup.
Signed-off-by: Sage Weil <sage@newdream.net>
Yehuda Sadeh [Tue, 15 Nov 2011 01:02:08 +0000 (17:02 -0800)]
rgw: don't log entries with bad utf8
Yehuda Sadeh [Mon, 14 Nov 2011 22:39:18 +0000 (14:39 -0800)]
rgw: adjust error code in swift copy failures
Yehuda Sadeh [Mon, 14 Nov 2011 21:55:09 +0000 (13:55 -0800)]
rgw: fix swift responses encoding
Josh Pieper [Fri, 11 Nov 2011 13:19:55 +0000 (08:19 -0500)]
rgw: Fix some merge problems uncovered by gcc warnings:
* a refactor in
e2100bce left the mod_ptr and unmod_ptr members set
incorrectly in RGWCopyObj::init_common
* a fix in
6752babd aggregated error returns, but then failed to do
anything with them
Signed-off-by: Josh Pieper <jjp@pobox.com>
Signed-off-by: Sage Weil <sage@newdream.net>
Josh Pieper [Fri, 11 Nov 2011 13:19:02 +0000 (08:19 -0500)]
Resolve gcc warnings.
These should have no functional changes:
* Check errors from functions that currently cannot return any
* Initialize variables that gcc can't determine will be initialized
in a following function call
* Remove unused variables
Signed-off-by: Josh Pieper <jjp@pobox.com>
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Mon, 14 Nov 2011 20:15:14 +0000 (12:15 -0800)]
osd: remove dead osd_max_opq code
This is no longer used as of a while ago!
Signed-off-by: Sage Weil <sage@newdream.net>
Josh Durgin [Mon, 14 Nov 2011 16:18:17 +0000 (08:18 -0800)]
workunits: rados python workunit should be executable
Signed-off-by: Josh Durgin <josh.durgin@dreamhost.com>
Sage Weil [Sun, 13 Nov 2011 22:18:19 +0000 (14:18 -0800)]
crush: send debug output to dout, not stdout/err
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sun, 13 Nov 2011 22:16:52 +0000 (14:16 -0800)]
test/run_cmd: use mkstemp instead of mkstemps
my box didn't have mkstemps
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sun, 13 Nov 2011 22:07:01 +0000 (14:07 -0800)]
ceph-authtool: fix clitests
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sat, 12 Nov 2011 23:17:29 +0000 (15:17 -0800)]
test_str_list: make sure ' ' and ', ' separaters work for str lists
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sat, 12 Nov 2011 22:55:39 +0000 (14:55 -0800)]
ceph-authtool: make error msg more helpful
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sat, 12 Nov 2011 22:55:28 +0000 (14:55 -0800)]
keyring: don't print auid if it is the default
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sat, 12 Nov 2011 22:55:06 +0000 (14:55 -0800)]
mon: implement 'fsid' command
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sat, 12 Nov 2011 22:19:26 +0000 (14:19 -0800)]
Merge branch 'stable'
Sage Weil [Sat, 12 Nov 2011 22:04:15 +0000 (14:04 -0800)]
mon: fix 'osd crush add ..' weight
This was changed to floating point in commit
3f67893 .
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sat, 12 Nov 2011 22:05:15 +0000 (14:05 -0800)]
osdmap: build_simple with normal osd/host/rack/pool hierarchy
This will be useful in the general case where the cluster is created with
an empty map and useful crush hierarchy.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sat, 12 Nov 2011 22:04:15 +0000 (14:04 -0800)]
mon: fix 'osd crush add ..' weight
This was changed to floating point in commit
3f67893 .
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sat, 12 Nov 2011 21:42:38 +0000 (13:42 -0800)]
vstart.sh: don't generate initial osdmap explicitly
This is simpler and exercises the monitors ability to start with a generic
osdmap and build it out as new osds are added to the cluster.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sat, 12 Nov 2011 21:41:59 +0000 (13:41 -0800)]
mon: make initial osdmap optional
If an initial osdmap is not provided, we generate an empty one. The user
add osds on their own after that.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sat, 12 Nov 2011 21:41:05 +0000 (13:41 -0800)]
osdmap: build_simple: create reasonable pools when numosd==0
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sat, 12 Nov 2011 21:16:30 +0000 (13:16 -0800)]
mon: add '--fsid foo' arg for setting generated monmap fsid
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sat, 12 Nov 2011 05:02:23 +0000 (21:02 -0800)]
mon: take '--fsid foo' arg with --mkfs
This will set the seed monmap's fsid. This is useful if the monmap is
dynamically generated (e.g., based on ceph.conf or --mon-host list).
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sat, 12 Nov 2011 05:03:09 +0000 (21:03 -0800)]
osd: fix warnings
osd/ReplicatedPG.cc: In member function 'virtual void ReplicatedPG::remove_watchers_and_notifies()':
osd/ReplicatedPG.cc:1167: warning: suggest a space before ';' or explicit braces around empty body in 'for' statement
osd/ReplicatedPG.cc:1176: warning: suggest a space before ';' or explicit braces around empty body in 'for' statement
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Sat, 12 Nov 2011 04:52:28 +0000 (20:52 -0800)]
monmaptool: fix clitests
Initial map is epoch 0. Modifications still bump epoch by one.
Signed-off-by: Sage Weil <sage.weil@dreamhost.com>
Sage Weil [Fri, 11 Nov 2011 23:38:24 +0000 (15:38 -0800)]
paxos: discard waiting_for_active events on reset
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 23:25:59 +0000 (15:25 -0800)]
monclient: use blank fsid (instead of epoch==0) for monmap checks
We can safely mkfs with an epoch=0 monmap as long as the fsid is set. And
that is what commit
f31825cee5300c708800a01a08201eef2bc03c0c changed.
Instead, use a zeroed fsid to tell if the monmap is valid/usable.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Sat, 12 Nov 2011 00:38:35 +0000 (16:38 -0800)]
use libuuid for fsid
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 22:59:38 +0000 (14:59 -0800)]
crush: grammer: allow '.' in name token
These are now in the generated crush maps, so it seems appropriate to
recompile them :).
Reported-by: Martin Mailand <martin@tuxadero.com>
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 22:54:41 +0000 (14:54 -0800)]
mon: fix seed monmap removal
Remove if we previous had no latest, not based on which map we now have.
It's possible we join when monmap epoch is something much larger than 1!
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 22:52:14 +0000 (14:52 -0800)]
mon: allow monitor to automagically join cluster
If a monitor starts up with the correct fsid and auth keys, it will now
add itself to the monmap (and subsequently try to join the quorum) if it
is not already in the monmap.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 20:52:24 +0000 (12:52 -0800)]
osd: pass monclient::init errors up the stack
Fixes crash like
ceph version
0.38-149-gbf254de (commit:
bf254de5cf8a17ce9467d166d87f3ab93170ae13 )
1: (ceph::BackTrace::BackTrace(int)+0x2d) [0x91d97b]
2: ./ceph-osd() [0xa05baa]
3: (()+0xef60) [0x7fb54c87ef60]
4: (std::_Rb_tree<unsigned int, unsigned int, std::_Identity<unsigned int>, std::less<unsigned int>, std::allocator<unsigned int> >::size() const+0xc) [0x8a4bc6]
5: (std::set<unsigned int, std::less<unsigned int>, std::allocator<unsigned int> >::size() const+0x18) [0x8a1d32]
6: (void encode<unsigned int>(std::set<unsigned int, std::less<unsigned int>, std::allocator<unsigned int> > const&, ceph::buffer::list&)+0x1c) [0x8a0311]
7: (MonClient::_reopen_session()+0x2c5) [0x89a425]
8: (MonClient::authenticate(double)+0x24f) [0x898da7]
9: (OSD::init()+0x112b) [0x807ca1]
10: (main()+0x2c09) [0x73e406]
11: (__libc_start_main()+0xfd) [0x7fb54b04ec4d]
12: ./ceph-osd() [0x73b499]
due to auth_supported being NULL.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 20:37:07 +0000 (12:37 -0800)]
mon: verify fsid during probe and election
This will keep mismatched fsids out of the same quorum.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 20:22:37 +0000 (12:22 -0800)]
mon: tolerate won election while active
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 20:22:22 +0000 (12:22 -0800)]
mon: clean up logic a bit
More explicit.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 20:22:09 +0000 (12:22 -0800)]
mon: only re bootstrap if monmap actually changes
If we go thru here just to update latest, that's fine; no need to restart
the bootstrap process.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 20:15:16 +0000 (12:15 -0800)]
paxos: fix off-by-one in share_state
We hit this on adding a new monitor to an existing cluster.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 20:05:01 +0000 (12:05 -0800)]
mon: fix monmap update
It's on the stack; update in place.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 20:02:52 +0000 (12:02 -0800)]
mon: properly process monmaps even when i have the latest
We may get the latest monmap when we are doing our probing, but we still
need to process it in update_from_paxos(). Consider get_latest_version()
in addition to the active map.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 19:55:34 +0000 (11:55 -0800)]
mon: fix up update_from_paxos() methods
Make sure they behave when the initial state is learned from paxos.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 19:40:20 +0000 (11:40 -0800)]
monmaptool: new maps get epoch 0
Just for consistency's sake.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 19:40:02 +0000 (11:40 -0800)]
mon: clean up mkfs seed data
And make sure the monmap/latest gets written properly.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 19:10:17 +0000 (11:10 -0800)]
mon: remove empty monstore dirs
This is sloppy, but it works well enough since we mkdir dirs as needed
too.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 19:01:50 +0000 (11:01 -0800)]
mon: create initial states after quorum is formed
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 18:45:27 +0000 (10:45 -0800)]
mon: stage mkfs seed info in mkfs/ dir
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 18:34:42 +0000 (10:34 -0800)]
mon: eliminate PaxosService::init()
update_from_paxos() is sufficient
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 18:19:33 +0000 (10:19 -0800)]
mon: include monmap dump in mon_status and quorum_status
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 18:15:23 +0000 (10:15 -0800)]
mon: pull initial monmap from monmap/latest OR mkfs/monmap
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Fri, 11 Nov 2011 18:05:36 +0000 (10:05 -0800)]
mon: take explicit initial monmap -or- generate one via MonClient
This will simplify bootstrapping a cluster via e.g. mon_host.
Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil [Thu, 10 Nov 2011 21:35:57 +0000 (13:35 -0800)]
test_filestore_idempotent: detect commit cycles due to non-idempotent ops
If we do a non-idempotent op and it does a commit itself, we don't see
fs->is_committed() true ever. Also count full commit cycles, and kill
ourselves after several of those have gone by.
Signed-off-by: Sage Weil <sage@newdream.net>