public inbox for cluster-cvs@sourceware.org
help / color / mirror / Atom feed
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-09-26 3:15 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-09-26 3:15 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-09-26 03:15:41
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c
Log message:
Bug 291521: Cluster mirror can become out-of-sync if nominal I/O overl...
Another touch-up for this bug.
Bad news:
Because a node can cache the state of a region indefinitely (especially for
blocks that are used alot - e.g. a journaling area of a file system), we must
deny writes to any region of the mirror that is not yet recovered. This is only
the case with cluster mirroring. This means poor performance of nominal I/O
during recovery - probably really bad performance. However, this is absolutely
necessary for mirror reliability.
Good news:
The time I spent coding different fixes for this bug weren't a complete waste.
I've been able to reuse some of that code to optimize the recovery process.
Now, rather than going through the mirror from front to back, it skips ahead to
recover regions that have pending writes. Bottom line: performance will be bad
during recovery, but it will be better than RHEL4.5.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.52&r2=1.1.2.53
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.14&r2=1.1.2.15
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.38&r2=1.1.2.39
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-10-03 19:02 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-10-03 19:02 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-10-03 19:02:52
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
Bug 316031: dm-mirror: incorrect order of mirror presuspend ops caus...
With kernel (dm-raid1.c) presuspend changes, we can now tell when
recovery has been shutdown and when we can allow blocked writes.
This should fix a hang issue when converting from one mirror type
to another.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.54&r2=1.1.2.55
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.40&r2=1.1.2.41
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-09-27 20:31 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-09-27 20:31 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-09-27 20:31:20
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c
Log message:
Bug 290821: cmirror write path appears deadlocked after recovery ...
In some device failure cases, regions must be marked 'out-of-sync' -
this was causing a following write to block because it thought the
region had not yet been recovered - when in fact, it had just been
put out-of-sync due to failing device.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.53&r2=1.1.2.54
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.15&r2=1.1.2.16
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.39&r2=1.1.2.40
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-09-21 20:07 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-09-21 20:07 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-09-21 20:07:37
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
dm-cmirror-xfr.h
Log message:
Bug 291521: Cluster mirror can become out-of-sync if nominal I/O overla...
It is insufficient to simply delay flush requests that have marks
pending to a recovering region. Although a collision between nominal
I/O and resync I/O can be avoided this way, the state of the region
changes from RH_NOSYNC to RH_CLEAN in the mean time. The machine
being delayed will think the region is still in the RH_NOSYNC state
and only write to the primary device... leaving the other mirror
devices out-of-sync.
We must delay writes to remotely recovering regions before the state
of the region is determined and cached in the region caching code...
The entry point for this already exists in 'is_remote_recovering'.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.51&r2=1.1.2.52
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.37&r2=1.1.2.38
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-xfr.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.6&r2=1.1.2.7
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-09-13 15:24 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-09-13 15:24 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-09-13 15:24:20
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
Bug 257881: Flush/recovery collision leads to deadlock after leg ...
The procedure for coordinating nominal I/O and recovery I/O, was to
either:
1) delay a flush which contained a mark to a region being recovered
2) skip over regions that are currently marked when assigning recovery
This bug has to do with the way #1 was implemented.
The following scenario would trigger it:
1) node1 is assigned recovery on region X
2) node1 also does a mark (write) on region Y
3) node2 attempts to mark region X
**) any flush issued here will delay waiting for recovery to complete on X
4) node1 needs to perform the flush before it can get on with completing
recovery - but it can't flush, so everyone is delayed *forever*.
The fix was to allow flushes from nodes that are not attempting to mark
regions that are being recovered. In the example above, node1 should be
allowed to complete the flush because it is not trying to write to the
same region that is being recovered. node2 would be correctly delayed.
Since node1 can complete the flush, it can also complete the recovery -
thus allowing things to proceed.
This bug only affects mirrors that are not in-sync and are doing I/O.
This bug can occur whether there are device/machine failures or not.
This bug is most easily reproduced with a number of mirrors, but would
be possible with just one.
I've also fixed up some debugging output so it is more consistent and
easier to follow the flow of events.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.50&r2=1.1.2.51
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.36&r2=1.1.2.37
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-07-11 16:18 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-07-11 16:18 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-07-11 16:18:03
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
dm-cmirror-xfr.h
Log message:
Bug 238629: dm-cmirror: Remote recovery conflict...
The kernel changes are now in place (marking/clearing the log
during writes to nosync regions) to allow nominal I/O to
region that have yet to be recovered.
Also moved around some debugging messages and removed
'is_remote_recovering()'. (is_remote_recovering is obviated by
the new mechanism for handling recovery/write ordering.)
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.48&r2=1.1.2.49
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.35&r2=1.1.2.36
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-xfr.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.5&r2=1.1.2.6
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-04-26 16:55 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-04-26 16:55 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL45
Changes by: jbrassow@sourceware.org 2007-04-26 17:55:51
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
Bug 238031: cluster mirrors not handling all recovery/write conflicts
Problem is that the kernel (main mirror code) does not do any marks/clears when
writing to a region before its recovery. So, it is not possible for the server
to detect a conflict. Basically, we must turn back on the
'is_remote_recovering' function and disallow any writes to regions that are OR
WILL BE recovering.
It's really going to cause some pain during writes while mirrors are re-syncing.
The better fix for the future is to have the writes always mark/clear the
regions - then we can again remove the 'is_remote_recovering' function.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.41.2.5&r2=1.1.2.41.2.6
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.26.2.7&r2=1.1.2.26.2.8
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-04-26 16:54 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-04-26 16:54 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-04-26 17:54:49
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
Bug 238031: cluster mirrors not handling all recovery/write conflicts
Problem is that the kernel (main mirror code) does not do any marks/clears when
writing to a region before its recovery. So, it is not possible for the server
to detect a conflict. Basically, we must turn back on the
'is_remote_recovering' function and disallow any writes to regions that are OR
WILL BE recovering.
It's really going to cause some pain during writes while mirrors are re-syncing.
The better fix for the future is to have the writes always mark/clear the
regions - then we can again remove the 'is_remote_recovering' function.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.46&r2=1.1.2.47
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.34&r2=1.1.2.35
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-04-24 20:10 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-04-24 20:10 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL45
Changes by: jbrassow@sourceware.org 2007-04-24 21:10:20
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
dm-cmirror-server.h dm-cmirror-xfr.h
Log message:
Bug 199433: NULL pointer dereference in cman:process_messages for cmirro...
- While this isn't a complete fix for 199433, it is most likely the
cause of the error. Cluster mirrors were steadily leaking memory
every time they were deactivated.
Bug 237028: cmirror recovery deadlock due to machine failure + primary l...
- If there is outstanding resync work remaining when the server gets
notice to suspend, delay for a moment to wait for it.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.41.2.4&r2=1.1.2.41.2.5
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.26.2.6&r2=1.1.2.26.2.7
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.h.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.2.8.1&r2=1.1.2.2.8.2
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-xfr.h.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.2.2.2&r2=1.1.2.2.2.3
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-04-24 20:08 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-04-24 20:08 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-04-24 21:08:57
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
dm-cmirror-server.h dm-cmirror-xfr.h
Log message:
Bug 199433: NULL pointer dereference in cman:process_messages for cmirro...
- While this isn't a complete fix for 199433, it is most likely the
cause of the error. Cluster mirrors were steadily leaking memory
every time they were deactivated.
Bug 237028: cmirror recovery deadlock due to machine failure + primary l...
- If there is outstanding resync work remaining when the server gets
notice to suspend, delay for a moment to wait for it.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.45&r2=1.1.2.46
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.33&r2=1.1.2.34
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.3&r2=1.1.2.4
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-xfr.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.4&r2=1.1.2.5
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-04-10 7:13 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-04-10 7:13 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL45
Changes by: jbrassow@sourceware.org 2007-04-10 08:13:15
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c dm-cmirror-server.h
Log message:
Bug 235686: Kernel BUG at dm_cmirror_server while recovering region
Several fixes have gone in to fix the handling of this bug:
1) During server relocation (which can happen due to machine failure
or normal mirror suspension), the server value could get set before
the client had a chance to clean-up. This caused the server to
become confused and issue a BUG().
2) perform a flush of the log before suspending. This ensures
that regions which are in-sync get correctly flushed to the disk
log. Without this, there will always be recovery work to be done
when a mirror starts up - even if it was properly in-sync during
shutdown.
3) clean-up memory used to record region users when a mirror is
shutdown. It was possible for some regions to be left over
(causing a memory leak) during certain fault scenarios.
4) properly initialize the state field (ru_rw) in the region
user structure when a mark occurs. Without the initialization,
it was sometimes possible for the region to be misinterpretted
as recovering instead of marked.
5) resolve and unhandled case in server_complete_resync_work
6) reset a variable in cluster_complete_resync_work. Failure
to do so was causing a retry to include the wrong value for
the completion of the resync work - confusing the server.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.41.2.3&r2=1.1.2.41.2.4
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.12.2.1&r2=1.1.2.12.2.2
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.26.2.4&r2=1.1.2.26.2.5
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.h.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.2&r2=1.1.2.2.8.1
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-04-10 7:12 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-04-10 7:12 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-04-10 08:12:24
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c dm-cmirror-server.h
Log message:
Bug 235686: Kernel BUG at dm_cmirror_server while recovering region
Several fixes have gone in to fix the handling of this bug:
1) During server relocation (which can happen due to machine failure
or normal mirror suspension), the server value could get set before
the client had a chance to clean-up. This caused the server to
become confused and issue a BUG().
2) perform a flush of the log before suspending. This ensures
that regions which are in-sync get correctly flushed to the disk
log. Without this, there will always be recovery work to be done
when a mirror starts up - even if it was properly in-sync during
shutdown.
3) clean-up memory used to record region users when a mirror is
shutdown. It was possible for some regions to be left over
(causing a memory leak) during certain fault scenarios.
4) properly initialize the state field (ru_rw) in the region
user structure when a mark occurs. Without the initialization,
it was sometimes possible for the region to be misinterpretted
as recovering instead of marked.
5) resolve and unhandled case in server_complete_resync_work
6) reset a variable in cluster_complete_resync_work. Failure
to do so was causing a retry to include the wrong value for
the completion of the resync work - confusing the server.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.44&r2=1.1.2.45
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.13&r2=1.1.2.14
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.30&r2=1.1.2.31
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.2&r2=1.1.2.3
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-04-05 21:33 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-04-05 21:33 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL45
Changes by: jbrassow@sourceware.org 2007-04-05 22:33:36
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
Bug 234918 Processed: NMI Watchdog detected LOCKUP while running proces...
Bug 217438: scrolling kernel requests to mark mirror regions
Item 1:
I needed to check for marked regions when getting resync work, not
just check for resyncing regions when a mark/flush happens.
Item 2:
There is a corner case that allows two calls to clear the same
region. The second does not need to be logged.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.41.2.2&r2=1.1.2.41.2.3
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.26.2.3&r2=1.1.2.26.2.4
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-04-05 21:32 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-04-05 21:32 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-04-05 22:32:26
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
Bug 234918 Processed: NMI Watchdog detected LOCKUP while running proces...
Bug 217438: scrolling kernel requests to mark mirror regions
Item 1:
I needed to check for marked regions when getting resync work, not
just check for resyncing regions when a mark/flush happens.
Item 2:
There is a corner case that allows two calls to clear the same
region. The second does not need to be logged.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.43&r2=1.1.2.44
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.29&r2=1.1.2.30
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-04-03 18:23 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-04-03 18:23 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL45
Changes by: jbrassow@sourceware.org 2007-04-03 19:23:01
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c dm-cmirror-xfr.h
Log message:
Bug 234539: multiple streams of I/O can cause system to lock up
This bug provoked an audit of the communications exchange, locking,
and memory allocations/stack usage.
Communication fixes include:
1) Added sequence numbers to ensure that replies from the server
correctly correspond to client requests. It was found that if
a client timed out waiting for a server to respond, it would send
the request again. However, the server may have simply been too
busy to respond in a timely fashion. It ends up responding to
both the original request and the resent request - causing the
client and server to become out-of-sync WRT log requests.
Locking fixes include:
1) A semaphore was being "up"ed twice in some cases, rendering
the lock impotent.
2) A spin lock controlling region status lists was being held
across blocking operations - sometimes causing deadlocks. The
spin lock was changed to a per-log lock, and some logging
operations were restructured to better suit the way locking
needed to be done. A side-effect of this fix is a 20%
improvement in write operations.
3) The log list protection lock needed to change from a spin lock
to a semaphore to allow blocking operations.
Memory allocation fixes include:
1) Wrong flags to kmalloc could cause deadlock. Use NOFS instead
of KERNEL.
2) Mempools needed more reserves for low memory conditions.
3) Server now allocates a communication structure instead of having
it on the stack. This reduces the likelyhood of stack corruption.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.41.2.1&r2=1.1.2.41.2.2
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.12&r2=1.1.2.12.2.1
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.26.2.1&r2=1.1.2.26.2.2
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-xfr.h.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.2.2.1&r2=1.1.2.2.2.2
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-04-03 18:21 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-04-03 18:21 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-04-03 19:21:10
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c dm-cmirror-xfr.h
Log message:
Bug 234539: multiple streams of I/O can cause system to lock up
This bug provoked an audit of the communications exchange, locking,
and memory allocations/stack usage.
Communication fixes include:
1) Added sequence numbers to ensure that replies from the server
correctly correspond to client requests. It was found that if
a client timed out waiting for a server to respond, it would send
the request again. However, the server may have simply been too
busy to respond in a timely fashion. It ends up responding to
both the original request and the resent request - causing the
client and server to become out-of-sync WRT log requests.
Locking fixes include:
1) A semaphore was being "up"ed twice in some cases, rendering
the lock impotent.
2) A spin lock controlling region status lists was being held
across blocking operations - sometimes causing deadlocks. The
spin lock was changed to a per-log lock, and some logging
operations were restructured to better suit the way locking
needed to be done. A side-effect of this fix is a 20%
improvement in write operations.
3) The log list protection lock needed to change from a spin lock
to a semaphore to allow blocking operations.
Memory allocation fixes include:
1) Wrong flags to kmalloc could cause deadlock. Use NOFS instead
of KERNEL.
2) Mempools needed more reserves for low memory conditions.
3) Server now allocates a communication structure instead of having
it on the stack. This reduces the likelyhood of stack corruption.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.42&r2=1.1.2.43
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.12&r2=1.1.2.13
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.27&r2=1.1.2.28
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-xfr.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.3&r2=1.1.2.4
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-03-22 22:34 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-03-22 22:34 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL45
Changes by: jbrassow@sourceware.org 2007-03-22 22:34:44
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
dm-cmirror-xfr.h
Log message:
Bug 233034: cmirror server failure/migration during GFS I/O causes metad...
Add sequence number to messages to ensure
that cmirror clients get the response they expect.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.41&r2=1.1.2.41.2.1
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.26&r2=1.1.2.26.2.1
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-xfr.h.diff?cvsroot=cluster&only_with_tag=RHEL45&r1=1.1.2.2&r2=1.1.2.2.2.1
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-03-22 22:22 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-03-22 22:22 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-03-22 22:21:59
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
dm-cmirror-xfr.h
Log message:
Bug 233034: cmirror server failure/migration during GFS I/O causes metad...
(Likely fixes other bugs as well.)
When a cmirror client timed out waiting for a response from the server,
it would send the request again. Sometimes, the server simply took to
long to get back to the client... it would then develop an off-by-one
error - responding to the first _and_ the second request. The client
could then be asking to mark a region, and recieve a response for
a previous request. This has the potential to cause many problems.
Sequence numbers have been added to fix the problem.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.41&r2=1.1.2.42
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.26&r2=1.1.2.27
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-xfr.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.2&r2=1.1.2.3
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-03-14 4:28 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-03-14 4:28 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-03-14 04:28:32
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
Bug 231230: leg failure on cmirrors causes devices to be stuck in SUSPE...
The problem here appears to be timeouts related to clvmd.
During failures under heavy load, clvmd commands (suspend/resume/
activate/deactivate) can take a long time. Clvmd assumes to quickly
that they have failed. This results in the fault handling being left
half done. Further calls to vgreduce (by hand or by dmeventd) will
not help because the _on-disk_ version of the meta-data is consistent -
that is, the faulty device has been removed.
The most significant change in this patch is the removal of the
'is_remote_recovering' function. This function was designed to check
if a remote node was recovering a region so that writes to the region
could be delayed. However, even with this function, it was possible
for a remote node to begin recovery on a region _after_ the function
was called, but before the write (mark request) took place. Because
of this, checking is done during the mark request stage - rendering
the call to 'is_remote_recovering' meaningless. Given the useless
nature of this function, it has been pulled. The benefits of its
removal are increased performance and much faster (more than an
order of magnitude) response during the mirror suspend process.
The faster suspend process leads to less clvmd timeouts and
reduced probability that bug 231230 will be triggered.
However, when a mirror device is reconfigured, the mirror sub-devices
are removed. This is done by activating them cluster-wide before
their removal. With high enough load during recovery, these operations
can still take a long time - even though they are linear devices.
This too has the potential for causing clvmd to timeout and trigger
bug 231230. There is no cluster logging fix for this issue. The
delay on the linear devices must be determined. A temporary
work-around would be to increase the timeout of clvmd (e.g. clvmd -t #).
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.40&r2=1.1.2.41
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.25&r2=1.1.2.26
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-02-26 17:38 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-02-26 17:38 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-02-26 17:38:06
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
add locking around the log list. There was a small window of opportunity
for the log server to look up a log in the list while another entry was
being deleted (bad for the server).
Bug 229715 Processed: cmirror panic in dm_cmirror:cluster_log_serverd
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.38&r2=1.1.2.39
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.24&r2=1.1.2.25
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-02-20 19:35 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-02-20 19:35 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-02-20 19:35:10
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-cman.c
dm-cmirror-cman.h dm-cmirror-server.c
Log message:
Bug 217895: lost election results from cmirror server cause mirror ...
There was a race happening as a result of simultaneous cman issued
'starts'. The client receives the start requests, but the server
processes them. So, it was possible for the server to reset the
event id/type while the client was trying to set them. This would
cause the next kcl_start_done command issued by the server to fail.
The bug can be interpretted many different ways depending on which
machine in the cluster you are looking at when it happens.
The fix was to have the client wait to set the event id/type until
it knows the server has completed the previous request.
This fix may resolve other bugs as well, but I will test them
individually.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.37&r2=1.1.2.38
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-cman.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.5&r2=1.1.2.6
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-cman.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.1&r2=1.1.2.2
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.22&r2=1.1.2.23
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-02-19 16:29 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-02-19 16:29 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-02-19 16:29:43
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
Bug 228070: DLM assertion when running GFS I/O during cmirror leg failure
When a log server drops out of the cluster, it ignores any requests -
forcing the clients to retry. Unfortunately, the clients never ran
another election - causing operations to stall. The server now replies
that it cannot handle the requests, which causes proper initiation of
elections.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.36&r2=1.1.2.37
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.21&r2=1.1.2.22
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-02-14 17:44 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-02-14 17:44 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-02-14 17:44:08
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c dm-cmirror-xfr.h
Log message:
Changes to fix the following bugs:
Bug 228104: greater than 2 legged cluster mirrors do not down ...
Bug 228056: lvconvert should give warning if we don't support ...
When converting from 3-way to 2-way mirror, the UUID for the mirror
stays the same. This creates conflicting entries in the cluster
logging code. I've added an additional identifier to allow for
unique identification in these cases.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.35&r2=1.1.2.36
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.10&r2=1.1.2.11
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.20&r2=1.1.2.21
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-xfr.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.1&r2=1.1.2.2
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-02-02 17:22 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-02-02 17:22 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-02-02 17:22:55
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
- Fix for bug #225337
Reset 'sync_search' if (lc->sync_search >= lc->region_count) &&
(lc->sync_count < lc->region_count). It indicates that a failure
during recovery has taken place, and we are likely able to handle
it.
Also, do not issue clear/mark region requests if it is already
known that the log device has failed.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.34&r2=1.1.2.35
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.19&r2=1.1.2.20
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2007-01-08 19:28 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2007-01-08 19:28 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2007-01-08 19:28:26
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
Now that the kernel is correctly handling sync state change, we can
remove the workaround in cmirror (keeping sync_count vs. decrementing
it).
Also moved some print statements to reduce console ouput.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.33&r2=1.1.2.34
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.18&r2=1.1.2.19
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-12-07 18:58 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-12-07 18:58 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2006-12-07 18:58:32
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
Get rid of a number of unnecessary messages, which spray to the console
during errors and cause the mirror reconfiguration to take a long time.
(Even been seen to cause machines to be fenced if the load is too great.)
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.32&r2=1.1.2.33
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.17&r2=1.1.2.18
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-09-05 17:50 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-09-05 17:50 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2006-09-05 17:50:11
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
- fix the bugs I've seen so far - mostly related to the recently added
ability to migrate the log server on suspension - that cause hangs
during combinations of create/delete/convert of mirrors
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.23&r2=1.1.2.24
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.16&r2=1.1.2.17
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-09-05 17:48 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-09-05 17:48 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4U4
Changes by: jbrassow@sourceware.org 2006-09-05 17:48:02
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
- fix the bugs I've seen so far - mostly related to the recently added
ability to migrate the log server on suspension - that cause hangs
during combinations of create/delete/convert of mirrors
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4U4&r1=1.1.2.19.2.4&r2=1.1.2.19.2.5
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4U4&r1=1.1.2.9.2.7&r2=1.1.2.9.2.8
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-07-27 23:11 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-07-27 23:11 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2006-07-27 23:11:55
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c
Log message:
- further tightening for 199826
We now refuse to read/write the disk log if we are suspended. We also
add BUG_ON(<suspended>) to operations that do I/O to the log device.
The reason for the BUG_ON() is that it is better to drop the machine
than to have it hang the cluster while it attempts to read/write from
a suspended device. That being said, it should now be impossible to
get to those functions which would perform I/O operations during
suspension.
I have still seen cases where the mirror will stall. However, I think
this is due to LVM (clvmd), because it happens when a mirror is created
while the log device is suspended - which must not happen. I've only
seen this when doing simultaneous create/convert/remove from all nodes
in the cluster.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.22&r2=1.1.2.23
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.9&r2=1.1.2.10
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.15&r2=1.1.2.16
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-07-27 23:11 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-07-27 23:11 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4U4
Changes by: jbrassow@sourceware.org 2006-07-27 23:10:58
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c
Log message:
- further tightening for 199826
We now refuse to read/write the disk log if we are suspended. We also
add BUG_ON(<suspended>) to operations that do I/O to the log device.
The reason for the BUG_ON() is that it is better to drop the machine
than to have it hang the cluster while it attempts to read/write from
a suspended device. That being said, it should now be impossible to
get to those functions which would perform I/O operations during
suspension.
I have still seen cases where the mirror will stall. However, I think
this is due to LVM (clvmd), because it happens when a mirror is created
while the log device is suspended - which must not happen. I've only
seen this when doing simultaneous create/convert/remove from all nodes
in the cluster.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4U4&r1=1.1.2.19.2.3&r2=1.1.2.19.2.4
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL4U4&r1=1.1.2.7.2.2&r2=1.1.2.7.2.3
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4U4&r1=1.1.2.9.2.6&r2=1.1.2.9.2.7
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-07-22 22:19 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-07-22 22:19 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: STABLE
Changes by: jbrassow@sourceware.org 2006-07-22 22:19:04
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c
Log message:
Fix for bug:
199862 - Suspending cluster mirrors can cause indefinite hang
And is part of a fix for:
199185 - 'lvconvert' fails to remove device-mapper devices ...
198555 - mirror log not getting cleared causes new mirror ...
And is likely to fix:
199334 - cmirror removal attempt hangs and caused locking ...
And will certainly help for:
199498
198821
194137
194125
199635
All of the above bugs will need to be reexamined when the packages
are rebuilt.
This fix allows the log server to migrate to other nodes during
suspension. This prevents the situation where the log server may
have it's devices suspended when it recieves a request. Trying to
fulfill a log request while devices are suspended will lead to an
indefinite hang, because I/O will not complete until the devices
are unsuspended.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=STABLE&r1=1.1.4.4&r2=1.1.4.5
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=STABLE&r1=1.1.4.3&r2=1.1.4.4
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=STABLE&r1=1.1.4.4&r2=1.1.4.5
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-07-22 22:19 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-07-22 22:19 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2006-07-22 22:19:34
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c
Log message:
Fix for bug:
199862 - Suspending cluster mirrors can cause indefinite hang
And is part of a fix for:
199185 - 'lvconvert' fails to remove device-mapper devices ...
198555 - mirror log not getting cleared causes new mirror ...
And is likely to fix:
199334 - cmirror removal attempt hangs and caused locking ...
And will certainly help for:
199498
198821
194137
194125
199635
All of the above bugs will need to be reexamined when the packages
are rebuilt.
This fix allows the log server to migrate to other nodes during
suspension. This prevents the situation where the log server may
have it's devices suspended when it recieves a request. Trying to
fulfill a log request while devices are suspended will lead to an
indefinite hang, because I/O will not complete until the devices
are unsuspended.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.21&r2=1.1.2.22
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.8&r2=1.1.2.9
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.13&r2=1.1.2.14
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-07-22 22:12 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-07-22 22:12 UTC (permalink / raw)
To: cluster-cvs
[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1586 bytes --]
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4U4
Changes by: jbrassow@sourceware.org 2006-07-22 22:12:33
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c
Log message:
Fix for bug:
199862 - Suspending cluster mirrors can cause indefinite hang
And is part of a fix for:
199185 – 'lvconvert' fails to remove device-mapper devices ...
198555 – mirror log not getting cleared causes new mirror ...
And is likely to fix:
199334 – cmirror removal attempt hangs and caused locking ...
And will certainly help for:
199498
198821
194137
194125
199635
All of the above bugs will need to be reexamined when the packages
are rebuilt.
This fix allows the log server to migrate to other nodes during
suspension. This prevents the situation where the log server may
have it's devices suspended when it recieves a request. Trying to
fulfill a log request while devices are suspended will lead to an
indefinite hang, because I/O will not complete until the devices
are unsuspended.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4U4&r1=1.1.2.19.2.2&r2=1.1.2.19.2.3
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL4U4&r1=1.1.2.7.2.1&r2=1.1.2.7.2.2
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4U4&r1=1.1.2.9.2.4&r2=1.1.2.9.2.5
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-06-29 19:49 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-06-29 19:49 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: STABLE
Changes by: jbrassow@sourceware.org 2006-06-29 19:49:32
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c
Log message:
- fix for bug 197263
The cluster mirror 'flush' logging function was still behaving the
way it used to before the changes to the kernel that allowed it to
receive the status of a log flush from 'flush'. This could result
in a indefinite suspension of the mirror on which a log device had
failed.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=STABLE&r1=1.1.4.2&r2=1.1.4.3
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=STABLE&r1=1.1.4.2&r2=1.1.4.3
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=STABLE&r1=1.1.4.2&r2=1.1.4.3
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-06-29 19:48 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-06-29 19:48 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2006-06-29 19:48:01
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c
Log message:
- fix for bug 197263
The cluster mirror 'flush' logging function was still behaving the
way it used to before the changes to the kernel that allowed it to
receive the status of a log flush from 'flush'. This could result
in a indefinite suspension of the mirror on which a log device had
failed.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.19&r2=1.1.2.20
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.7&r2=1.1.2.8
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.11&r2=1.1.2.12
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-06-29 19:46 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-06-29 19:46 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4U4
Changes by: jbrassow@sourceware.org 2006-06-29 19:46:37
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c
Log message:
- fix for bug 197263
The cluster mirror 'flush' logging function was still behaving the
way it used to before the changes to the kernel that allowed it to
receive the status of a log flush from 'flush'. This could result
in a indefinite suspension of the mirror on which a log device had
failed.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4U4&r1=1.1.2.19&r2=1.1.2.19.2.1
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL4U4&r1=1.1.2.7&r2=1.1.2.7.2.1
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4U4&r1=1.1.2.9.2.2&r2=1.1.2.9.2.3
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-06-27 20:19 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-06-27 20:19 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: STABLE
Changes by: jbrassow@sourceware.org 2006-06-27 20:19:53
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c dm-cmirror-xfr.h
dm-log.h
Log message:
- bring logging functions inline with upstream API
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=STABLE&r1=1.1.4.1&r2=1.1.4.2
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=STABLE&r1=1.1.4.1&r2=1.1.4.2
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=STABLE&r1=1.1.4.1&r2=1.1.4.2
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-xfr.h.diff?cvsroot=cluster&only_with_tag=STABLE&r1=1.1.4.1&r2=1.1.4.2
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-log.h.diff?cvsroot=cluster&only_with_tag=STABLE&r1=1.1.4.1&r2=1.1.4.2
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-06-15 19:48 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-06-15 19:48 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2006-06-15 19:48:00
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-cman.c
Log message:
- pull some more unnecessary prints (and test cvs commit)
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.18&r2=1.1.2.19
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-cman.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.4&r2=1.1.2.5
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-06-15 19:34 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-06-15 19:34 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2006-06-15 19:34:41
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-cman.c
Log message:
- last typo and a message clean-up
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.17&r2=1.1.2.18
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-cman.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.3&r2=1.1.2.4
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-06-13 16:26 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-06-13 16:26 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2006-06-13 16:26:15
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-cman.c
dm-cmirror-server.c
Log message:
- downgrade some messages to INFO
- serialize client requests to prevent getting wrong response
- properly clear regions before shutting down a mirror
- do not allocate memory while holding a spin lock
- do not set sync bit multiple times if network failures occur
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.14&r2=1.1.2.15
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-cman.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.2&r2=1.1.2.3
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.7&r2=1.1.2.8
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-06-02 22:04 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-06-02 22:04 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2006-06-02 22:04:51
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
- Fix for bugs: 193596 193706
- Potential fix for: 193736 193902 (these fix message output change only)
193596: node recovery can cause cluster mirror copying to halt
Before we had the ability to detect whether or not another member node was recovering, I made sure that only the server would recover the mirror. However, this meant that other nodes would recieve the signal not to recover. So, in this case, when the server failed/left, recovery would stop.
Now that we have the ability to detect remote recovery, I have remove the stipulation that only the server can recover. This has fixed the problem for me.
193706: cluster mirror copy starts over when node is recovered
This was a problem were I was making the complete_resync_work function handle what should have been in the clear_region function for recovery scenarios. The result was that although the clear_region was happening, that bitmap was not properly being written.
193736: deactivating an active cmirror will can cause 1000s of d-m messages
Yeah, I think I put that in there for my own debugging...
Changed DMERR -> DMINFO.
193902: mirror volume removal causes "A node has left the cluster" message
Proposing "A cluster mirror log member has left."
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.13&r2=1.1.2.14
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.6&r2=1.1.2.7
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-05-15 16:26 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-05-15 16:26 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2006-05-15 16:26:56
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
- take account of the fact that the sync status can now change
orthogonally to machine liveness (e.g. in response to disk
failures).
We can no longer assume that the only way for a cluster mirror
to go "out-of-sync" is for a machine to fail/leave.
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.12&r2=1.1.2.13
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.5&r2=1.1.2.6
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-02-08 21:59 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-02-08 21:59 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2006-02-08 21:59:33
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-cman.c
Log message:
- printk -> DMxxx
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.11&r2=1.1.2.12
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-cman.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.1&r2=1.1.2.2
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-02-08 16:03 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-02-08 16:03 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2006-02-08 16:03:17
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c
Log message:
- endianness fixes
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.10&r2=1.1.2.11
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.5&r2=1.1.2.6
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.4&r2=1.1.2.5
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-02-06 23:04 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-02-06 23:04 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2006-02-06 23:04:01
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-log.h
Log message:
- rtn type for get_region_size is uint32_t
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.9&r2=1.1.2.10
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-log.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.3&r2=1.1.2.4
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-02-06 22:58 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-02-06 22:58 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2006-02-06 22:58:15
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-server.c
Log message:
- remove some unneeded prints
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.8&r2=1.1.2.9
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.3&r2=1.1.2.4
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2006-02-06 18:14 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2006-02-06 18:14 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2006-02-06 18:14:20
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
Log message:
- changes to table line parsing to go along with what is needed by
user-space tools
Patches:
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.7&r2=1.1.2.8
http://sourceware.org/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.4&r2=1.1.2.5
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2005-10-27 21:27 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2005-10-27 21:27 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2005-10-27 21:27:18
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-log.h
Log message:
- fix some build errors
Patches:
http://sources.redhat.com/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.6&r2=1.1.2.7
http://sources.redhat.com/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-log.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.2&r2=1.1.2.3
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2005-10-27 11:03 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2005-10-27 11:03 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2005-10-27 11:03:23
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-log.h
Log message:
- first set of changes to bring cluster mirroring back in sync
Patches:
http://sources.redhat.com/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.4&r2=1.1.2.5
http://sources.redhat.com/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.3&r2=1.1.2.4
http://sources.redhat.com/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-log.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.1&r2=1.1.2.2
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2005-08-11 18:26 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2005-08-11 18:26 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2005-08-11 18:26:19
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c dm-cmirror-server.h
Log message:
- fix problem associating a log to a unique id.
Patches:
http://sources.redhat.com/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.2&r2=1.1.2.3
http://sources.redhat.com/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.2&r2=1.1.2.3
http://sources.redhat.com/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.2&r2=1.1.2.3
http://sources.redhat.com/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.1&r2=1.1.2.2
^ permalink raw reply [flat|nested] 51+ messages in thread
* cluster/cmirror-kernel/src dm-cmirror-client.c ...
@ 2005-07-28 22:53 jbrassow
0 siblings, 0 replies; 51+ messages in thread
From: jbrassow @ 2005-07-28 22:53 UTC (permalink / raw)
To: cluster-cvs
CVSROOT: /cvs/cluster
Module name: cluster
Branch: RHEL4
Changes by: jbrassow@sourceware.org 2005-07-28 22:53:24
Modified files:
cmirror-kernel/src: dm-cmirror-client.c dm-cmirror-common.h
dm-cmirror-server.c
Log message:
- fix cmirror bugs
- Only allow one machine to perform resync work
There must be a race in there somewhere when allowing multiple
simultaneous recoveries... This side-steps that issue. Besides
one machine moving the heads is probably better than all.
- when a resync is done, we must mark it in the clean list as well
as the sync list. Otherwise, the log will always tell us there
is recovery to be done - even if there isn't.
- Only suspend the server on cluster events - not dm suspends. If
we flush the clients before returning from the suspend, the server
will do nothing anyway.
- fix typo
- clean up logging
- take the code that connects to cman out of the init function and
do it on first mirror activation. This way, you don't have to
have cman running in order to load the module... Makes mirror
activation a bit slower, but speeds up module loading.
Patches:
http://sources.redhat.com/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-client.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.1&r2=1.1.2.2
http://sources.redhat.com/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-common.h.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.1&r2=1.1.2.2
http://sources.redhat.com/cgi-bin/cvsweb.cgi/cluster/cmirror-kernel/src/dm-cmirror-server.c.diff?cvsroot=cluster&only_with_tag=RHEL4&r1=1.1.2.1&r2=1.1.2.2
^ permalink raw reply [flat|nested] 51+ messages in thread
end of thread, other threads:[~2007-10-03 19:02 UTC | newest]
Thread overview: 51+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2007-09-26 3:15 cluster/cmirror-kernel/src dm-cmirror-client.c jbrassow
-- strict thread matches above, loose matches on Subject: below --
2007-10-03 19:02 jbrassow
2007-09-27 20:31 jbrassow
2007-09-21 20:07 jbrassow
2007-09-13 15:24 jbrassow
2007-07-11 16:18 jbrassow
2007-04-26 16:55 jbrassow
2007-04-26 16:54 jbrassow
2007-04-24 20:10 jbrassow
2007-04-24 20:08 jbrassow
2007-04-10 7:13 jbrassow
2007-04-10 7:12 jbrassow
2007-04-05 21:33 jbrassow
2007-04-05 21:32 jbrassow
2007-04-03 18:23 jbrassow
2007-04-03 18:21 jbrassow
2007-03-22 22:34 jbrassow
2007-03-22 22:22 jbrassow
2007-03-14 4:28 jbrassow
2007-02-26 17:38 jbrassow
2007-02-20 19:35 jbrassow
2007-02-19 16:29 jbrassow
2007-02-14 17:44 jbrassow
2007-02-02 17:22 jbrassow
2007-01-08 19:28 jbrassow
2006-12-07 18:58 jbrassow
2006-09-05 17:50 jbrassow
2006-09-05 17:48 jbrassow
2006-07-27 23:11 jbrassow
2006-07-27 23:11 jbrassow
2006-07-22 22:19 jbrassow
2006-07-22 22:19 jbrassow
2006-07-22 22:12 jbrassow
2006-06-29 19:49 jbrassow
2006-06-29 19:48 jbrassow
2006-06-29 19:46 jbrassow
2006-06-27 20:19 jbrassow
2006-06-15 19:48 jbrassow
2006-06-15 19:34 jbrassow
2006-06-13 16:26 jbrassow
2006-06-02 22:04 jbrassow
2006-05-15 16:26 jbrassow
2006-02-08 21:59 jbrassow
2006-02-08 16:03 jbrassow
2006-02-06 23:04 jbrassow
2006-02-06 22:58 jbrassow
2006-02-06 18:14 jbrassow
2005-10-27 21:27 jbrassow
2005-10-27 11:03 jbrassow
2005-08-11 18:26 jbrassow
2005-07-28 22:53 jbrassow
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).