1
0
mirror of git://sourceware.org/git/lvm2.git synced 2024-12-25 01:34:38 +03:00
Commit Graph

3961 Commits

Author SHA1 Message Date
Milan Broz
ff442c1e28 Propagate commit and revert metadata event to other nodes in cluster.
This patch tries to correctly track changes in lvmcache related to commit/revert.

For vg_commit: if there is cached precommitted metadata, after successfull commit
these metadata must be tracked as committed.

For vg_revert: remote nodes must drop precommitted metadata and its flag in lvmcache.

(N.B. Patch do not touch LV locks here in any way.)

All this machinery is needed to properly solve remote node cache invalidaton which
cause several problems recently observed.
2010-01-05 16:09:33 +00:00
Milan Broz
5cedaa972a Proper mask lock mode for vg lock.
Lock mode is int masked by LCK_TYPE_MASK, always.

Patch also remove uneccessary masking lock flag on sender side,
if masking is needed, it is don on client side already.
2010-01-05 16:07:56 +00:00
Milan Broz
2c8f647e0c Add possibility to handle precommitted metadata in lvmcache.
- Add drop_precommitted flag to force drop precommitted metadata
 - add lvmcache_commit_metadata() which upgrades precommitted metadata in cache

No functional change in this patch - just preparation for following change.
2010-01-05 16:06:42 +00:00
Milan Broz
3be5edae0e Move processing of VG locks to separate function (similar to LV locks).
And print some debugging info.

No functional change in this patch.
2010-01-05 16:05:12 +00:00
Milan Broz
1078f24d5a Properly decode flags even for VG locks.
And decode flags in humar readable form in client.
And clean some trailing whitespaces.

No functional change in this patch (only debugging messages changed).
2010-01-05 16:03:37 +00:00
Milan Broz
c02cc5321a Do not set precommitted flag in cache when precommitted metadata does not exist.
The use_precommitted flag indicates, that we want to use precommitted metadata
(used in suspend call to preload table with precommitted data).

But if there are no such data, committed metadata are read but the cache
still contains that precommitted flag.

(The problem is that later possible drop_metadata call will not invalidate
device in cache.)

The wrong precommitted state is stored in on remote nodes during normal
suspend/resume cycle _without_ vg_write/commit.

Use the PRECOMMITTED status flag here instead (which is always set if using
precommited metadata here).
2010-01-05 16:01:22 +00:00
Milan Broz
e60a7ff1d0 Resume volumes in reverse order to preserve memlock pairing.
If renaming snapshot with virtual origin, the origin is renamed too.
But the code must resume LVs in reverse order to properly
pair memlock (in cluster locking).

(The resume of snapshot resumes origin too and later resume
is ignored otherwise.)
2010-01-05 15:58:11 +00:00
Milan Broz
91de7311d7 Fix previous vgcreate commit to not call unpaired unlock. 2009-12-28 18:34:45 +00:00
Milan Broz
e357c9fc0c Explicitly use non-clustered vgcreate in test.
(So the tests can run under cluster locking and do not require
cluster mirror or snapshots.)

Add vgscan before block device readahead change
(flush long running process - clvmd - dev cache.)
2009-12-28 18:33:04 +00:00
Milan Broz
ec4daa5936 Drop metadata cache after device was autorepaired and removed from VG.
All long running processes must reload metadata when some
device becomes orphan after repair.
2009-12-18 12:45:41 +00:00
Milan Broz
2743fb9771 Remove missing flag if PV reappeared and is empty.
When PV device reappears with old metadata, it is
always updated to new version byt atutomatic metadata
repair.

Remove missing flag if device is empty.

If device contains allocated extents, issue warning that
user must remove volumes and re-add this PV before
manipulating with this volume.

This partially solves bug 547842 when one PV (log) is failed,
dmeventd removes that device and later this device reappears and
is wrongly added into VG marked missing.
2009-12-18 12:44:20 +00:00
Petr Rockai
88c4802bdc Revert another unintended change that snuck in. 2009-12-17 15:59:53 +00:00
Petr Rockai
a47e422715 Fix removal of multiple devices from a mirror (+ regression test). 2009-12-17 15:38:29 +00:00
Petr Rockai
3a9ab631e5 Also clean up MISSING devices in --removemissing --force in vgreduce. 2009-12-17 13:54:46 +00:00
Petr Rockai
9ca5118a68 Revert unintended change that slipped in with last checkin. 2009-12-16 19:26:20 +00:00
Petr Rockai
d6053080f8 #define an INTERNAL_ERROR macro and use it throughout LVM. 2009-12-16 19:22:11 +00:00
zkabelac
b1f8076da0 Cleanup returns for void functions. 2009-12-11 13:16:37 +00:00
zkabelac
abc760c8e4 Destroy allocated mempool in _vg_read_orphans() error path. 2009-12-11 13:14:44 +00:00
zkabelac
0331a34d58 Fix unlocking vg in some pvresize and toollib error paths. 2009-12-11 13:11:56 +00:00
zkabelac
98af0ba143 Fix coredump and memory leak for 'dmsetup help -c' 2009-12-11 13:04:30 +00:00
Jonathan Earl Brassow
157810ea3d s/=/==/ Typo was causing sub test to always return success. 2009-12-10 22:06:15 +00:00
Milan Broz
448e4251f8 Call explicitly suspend for temporary mirror layer.
The memlock_inc() fix is wrong, memlock count is not
propagated to long living process (clvmd) and just
it underflow there.
Also suspend is needed to pre-load precommited metadata
on other nodes (remapping to error taget in this case).

With explicit suspend we generate lock request and code
can update memlock count.

(Infinitely "locked" memory caused that fs_unlock() was not
called properly and on cluster nodes remains
old links in /dev/mapper for not active devices.)

(N.B. failing of suspend call here is not handled as fatal
error - the LV is going to be removed later anyway.)
2009-12-09 19:53:39 +00:00
Milan Broz
81ce20d616 Use more descriptive variable name for temporary layer lv. 2009-12-09 19:43:39 +00:00
Milan Broz
3e9ab41e76 Fix missing include. 2009-12-09 19:30:56 +00:00
Milan Broz
a3c8fbc6b6 Allow manipulation with precommited metadata even when a PV is missing.
The new recovery code first tries to repair LV and then removes failed PV
from VG. It means that during operation there can be VG with PV missing,
and vg_read code handles it like not consistent VG.

We already allows returning "inconsistent" commited metadata,
for mirror repair we need this for precommited too.
(The suspend call prepares precommited metadata to inactive table on
other cluster nodes.)

"Inconsistent" here means - correct metadata, just with some metadata areas
not found (obviously on missing or failed PVs).
2009-12-09 19:29:04 +00:00
Milan Broz
e9dc6880af Add memlock information to do_lock_lv debug output.
This helps a lot to detect that something strange happened.
2009-12-09 19:01:27 +00:00
Milan Broz
25fb2c4fb6 Never ever use distributed lock for LV in non-clustered VG.
The LV locks make sense only for clustered LVs.

Properly check cluster flag and never issue cluster lock here.

There are several places in code, where it is already checked, this
patch add this check to all needed calls.

In previous code the lock behaviour was inconsistent,
for example, the pre/post callback can take lock even for local volume,
but deactivate call do not released this lock and it remains held forever.

The local LV lock request now just let run the underlying activation code
on local node, the same process like in local locking.

(Again, this is important for new mirror repair calls, here for local
mirrors but with cluster locking enabled.)
2009-12-09 19:00:16 +00:00
Milan Broz
97d28ae4d7 Allow implicit lock conversion for pre/post callbacks.
This is unnoticed regression from commit 31672ff60e

The pre/post callback need to convert lock always, local node
is going to modify metadata in this case, it it fails conversion,
the call is ignored.

Also it fixes bug when the lock is not yet held, we cannot set LKF_CONVERT
in this case, it will fail because this lock do not exist.

Note that the automatic conversion is still disabled in activate
call, so the original fix (reactivation of exlusive LV) should
be still in place.
2009-12-09 18:55:53 +00:00
Milan Broz
b5bad15dae Allow implicit "convert" to the same lock mode.
(Code already not fail if unlocking not locked resource.)

This is needed in pre/post lock_lv call, where we can
request the same lock on local node becuase of suspend call.
2009-12-09 18:45:12 +00:00
Milan Broz
3098ef4a6e Get rid of magic masks in cluster locking code - clvmd part.
- do_command and lock_vg expect flags (no change here)

Bug fixes:
- lock_vg should check for NONBLOCK on lock_cmd, flags have this bit masked-out

- do_pre/post_command expect do not mask flag at all, this causes that
the code inside is never run! (see following patches, these functions
expect plain command without flags)
2009-12-09 18:42:02 +00:00
Milan Broz
e354b1210e Get rid of magic masks in cluster locking code.
Patch should not cause any problems, only real change is
removing LCK_LOCAL bit from lock type flag, it is never used there.
(LCK_LOCAL is part arg[1] bits anyway.)
2009-12-09 18:28:27 +00:00
Milan Broz
895b6ff719 Get rid of hardcoded 0xffdf cluster lock flag.
There is hidded change - the upper flags (0xffff0000)
and now not cleared, but there are unused anyway.
2009-12-09 18:16:38 +00:00
Milan Broz
0a22655ff3 Remove newly created log volume if initial deactivation fails.
If there is problem deactivate LV and
_init_mirror_log is called with remove_on_failure = 1,
remove the newly created log LV from metadata.

(This can happen if there is active device with the same name
but different UUID.)

The main reason for this "workaround" patch is to
 - do not keep _mlog volume in metadata, so user can repeat the action
 - print better error message describing the real problem

# lvcreate -m 2 -n lv1 -l 1 --nosync vg_bar
  WARNING: New mirror won't be synchronised. Don't read what you didn't write!
  /dev/vg_bar/lv1_mlog: not found: device not cleared
  Aborting. Failed to wipe mirror log.
  Error locking on node bar-01: Input/output error
  Unable to deactivate mirror log LV. Manual intervention required.
  Failed to create mirror log.

# lvcreate -m 2 -n lv1 -l 1 --nosync vg_bar
  WARNING: New mirror won't be synchronised. Don't read what you didn't write!
  Aborting. Unable to deactivate mirror log.
  Failed to initialise mirror log.
2009-12-09 18:09:52 +00:00
Dave Wysochanski
8def378b5e Fix activated/deactivated log_verbose message
I see "Deactivated" message when I activate and "Activated" message when
I deactivate.  The code uses "activate" as boolean but it can be any one
of the enum values from CHANGE_AY, CHANGE_AN, CHANGE_AE, etc.

Signed-off-by: Malahal Naineni <malahal@us.ibm.com>
Acked-by: Dave Wysochanski <dwysocha@redhat.com>
2009-12-07 19:32:28 +00:00
Peter Rajnoha
7ab8b44b6c Disable udev rules on change event with DISK_RO=1.
There's a new change udev event generated since kernel 2.6.32 that
notifies userspace about a change in read-only attribute for block
devices (with DISK_RO=1 environment variable set).

We need to detect this and disable the rule application so the
meaning of this change event is not interchanged with the regular
change event used while resuming/renaming DM devices.

If there's anybody awaiting this notification in foreign rules,
he can still check for this env var and do the appropriate actions
separately.
2009-12-07 12:03:47 +00:00
Dave Wysochanski
335fe09d70 Update a few more uint64_t's related to the 64-bit status change.
At this point they probably do not matter but going forward they
may - depends on future patches for replicator, etc.  I think
these probably got missed because they were 'flags' so I changed
the name to 'status' to be consistent.  So the on-disk
things 'flags' and the in structure 'status' (bits).
NOTE: WHATS_NEW already has entry for this in current release.

Signed-off-by: Dave Wysochanski <dwysocha@redhat.com>
Acked-by: Mike Snitzer <snitzer@redhat.com>
2009-12-04 17:48:32 +00:00
Peter Rajnoha
523a7ed7d3 WHATS_NEW for previous commit. 2009-12-04 14:26:22 +00:00
Peter Rajnoha
f69e8e3606 Give better message for pvmove when all data on source PV is skipped. 2009-12-04 14:03:18 +00:00
Milan Broz
3516797b49 Fix memory lock imbalance in lv_suspend if already suspended.
pvmove suspends all moved LVs + pvmoveX mirrored LV itself.

This suspends even underlying pvmoveX and following explicit
suspend call is just noop.

But in resume the pvmoveX volume is no longer underlying
device for moved LVs, so it performs full resume with memlock
decrease.

Code must call memlock_inc() if suspend is requested, volume
is already suspended and error is not requested.
2009-12-03 19:23:40 +00:00
Milan Broz
1683e94f0c Fix pvmove test mode to not fail and do not poll.
Test mode should not fail nor try to poll non-existent devices.
2009-12-03 19:22:24 +00:00
Milan Broz
1d249195a4 Print error if VG already exist.
This test have to be moved because of new vg read error handling.
2009-12-03 19:20:48 +00:00
Milan Broz
2f7ee08288 Fix tools to report error when stopped by user.
(And do not produce internal error message.)
2009-12-03 19:18:33 +00:00
zkabelac
e0772fb301 minor indent change 2009-12-03 10:01:30 +00:00
zkabelac
78bdbce2fd skip cast from (void*) 2009-12-03 09:59:54 +00:00
zkabelac
5a5fed28fa minor whitespace indentation 2009-12-03 09:58:30 +00:00
Dave Wysochanski
5818b6efb8 Add tests to check for readahead value in lvcreate. 2009-12-03 01:48:05 +00:00
Dave Wysochanski
53a66492e8 Fix setting of readahead in lvcreate.
The default comes from the configuration settings, with possible
commandline override.
2009-12-03 01:47:33 +00:00
Milan Broz
b17ddc26c0 Fix memory leak in lv_info_by_lvid
The lv_from_lvid calls internally vg_read(),
we must release vg structure afterwards.

Code is called only from clvmd.
2009-12-01 19:10:23 +00:00
Petr Rockai
b7e19d2fe3 If aborting due to an internal error, always print the message causing this. 2009-12-01 13:54:27 +00:00
Petr Rockai
d8130a99b4 Optionally abort on internal errors (and leverage this option in the
testsuite). (This is showing a problem in the pvmove test for me, so I expect
the tests to start failing -- this needs to be fixed separately though.)
2009-11-30 17:17:11 +00:00