[DRBD-user] maximal mount count reached

ha at buglecreek.com ha at buglecreek.com
Wed Jan 28 21:10:58 CET 2009


Thanks for the information.  Just want to clarify a few points to make
sure I understand.  In the following case - Node A Primary and Node B
Secondary. See below:

On Wed, 28 Jan 2009 11:04:18 +0100, "Lars Ellenberg"
<lars.ellenberg at linbit.com> said:
> Do not hijack threads.
> 
> On Tue, Jan 27, 2009 at 04:30:44PM -0700, ha at buglecreek.com wrote:
> > I am in the process of upgrading the OS on a 2 node cluster that is
> > running heartbeat-2.1.4-2.1 (R1 style), drbd-8.0.14, drbdlinks.  I have
> > rebuilt one node and while switching back and forth between nodes while
> > testing I noticed the following log entries:
> > 
> > INFO: Running start for /dev/drbd0 on /shared
> > kernel: EXT3-fs warning: maximal mount count reached, running e2fsck is
> > recommended
> > kernel: EXT3 FS on drbd0, internal journal
> > 
> > When heartbeat mounts the drbd device I am getting the maximal mount
> > count warning.  I'm not sure how to handle this on a drbd device.  I
> > have a simple setup with a separate disk on each node dedicated to drbd.
> >  No lvm, raid etc.  It just gets mounted to /shared by heartbeat. I'm
> > getting this on both nodes. Can you do a fsck on the backing device
> > (/dev/hda1) on the node that is not primary without causing problems? 
> 
> No. Do not bypass drbd.
> 
> > Can anyone suggest a way to handle this?  
> 
> either forget about the fsck.
> if the warnings annoy you,
> disable the mount count and mount time counters,
> see tune2fs -c and -i (use on drbd0!)

Can I run tune2fs -c -1 -i 0 /dev/drbd0 on Node B since it is not
mounted to /shared.  Or do I have to unmount /shared on Node A before I
can run the tune2fs?  In other words, does /dev/drbd0 have to be NOT
mounted on either node?

> 
> or fsck the DRBD.
> -> umount /shared, fsck drbd0, mount /shared.

Same thing here can I fsck /dev/drbd0 on Node B or do I have to make
sure that the /dev/drbd0 device is not mounted on either node.

> 
> as you say you use heartbeat in haresources mode,
> that should be straight forward.
> 
> if you where in CRM mode, you'd need to set the target role of the
> Filesystem to stopped temporarily, or set it to unmanaged.
> 
> -- 
> : Lars Ellenberg
> : LINBIT | Your Way to High Availability
> : DRBD/HA support and consulting http://www.linbit.com
> 
> DRBD® and LINBIT® are registered trademarks of LINBIT, Austria.
> __
> please don't Cc me, but send to list   --   I'm subscribed
> _______________________________________________
> drbd-user mailing list
> drbd-user at lists.linbit.com
> http://lists.linbit.com/mailman/listinfo/drbd-user



More information about the drbd-user mailing list