[Drbd-dev] DRBD8: unconfigured devices and scheduling after_state_ch problem

[Drbd-dev] DRBD8: unconfigured devices and scheduling after_state_ch problem

[Montrose, Ernest]

[-- Attachment #1.1: Type: text/plain, Size: 518 bytes --]

Hi all,
We have been seeing a problem where "unconfigured" devices e.g devices
that are Standalone and Diskless with sequential sync order schedule
after_state_ch().  The problem is that there isn't a worker thread in
this case.  So we observed two side effects. One all these queued work
make shutting down the node take a long time as unloading the driver
takes time to clean up. Two we can potentially have a memory exhaustion
issue. 
 
Included here is a patch to help fix the problem.
 
Thanks.
EM--

[-- Attachment #1.2: Type: text/html, Size: 4749 bytes --]

[-- Attachment #2: after_isp.patch --]
[-- Type: application/octet-stream, Size: 548 bytes --]

Index: drbd/drbd_worker.c
===================================================================
--- drbd/drbd_worker.c	(revision 19381)
+++ drbd/drbd_worker.c	(working copy)
@@ -797,7 +797,8 @@
 	int i, rv = 0;
 
 	for (i=0; i < minor_count; i++) {
-		if( !(odev = minor_to_mdev(i)) ) continue;
+		if( !(odev = minor_to_mdev(i)) ||
+                  (odev->state.conn == StandAlone && odev->state.disk == Diskless) ) continue;
 		if (! _drbd_may_sync_now(odev)) {
 			rv |= ( _drbd_set_state(_NS(odev,aftr_isp,1),
 						ChgStateHard|ScheduleAfter)

Re: [Drbd-dev] DRBD8: unconfigured devices and scheduling after_state_ch problem

[Philipp Reisner]

On Tuesday 16 October 2007 23:55:49 Montrose, Ernest wrote:
> Hi all,
> We have been seeing a problem where "unconfigured" devices e.g devices
> that are Standalone and Diskless with sequential sync order schedule
> after_state_ch().  The problem is that there isn't a worker thread in
> this case.  So we observed two side effects. One all these queued work
> make shutting down the node take a long time as unloading the driver
> takes time to clean up. Two we can potentially have a memory exhaustion
> issue.
>
> Included here is a patch to help fix the problem.
>

Right, I applied this...

-Phil
-- 
: Dipl-Ing Philipp Reisner                      Tel +43-1-8178292-50 :
: LINBIT Information Technologies GmbH          Fax +43-1-8178292-82 :
: Vivenotgasse 48, 1120 Vienna, Austria        http://www.linbit.com :