[Ocfs2-users] OCFS processes active after a umount [SEC=UNOFFICIAL]

[Mark Schloss]

Hello Sunil,

Thanks for the reply. The version is 1.2.9-1 and kernel is
2.6.18-92.1.1.el5.

We have managed to reproduce the problem and it appears to be related to
multipathing. We recently moved to multipathing the OCFS volumes. On the
weekend we tested removing one path to the OCFS volumes and all was OK.
When the path was restored the box still showed it as being unavailable.
We found this to be caused by the multipath daemon not running as it had
not been set to start automatically. After the daemon was started both
paths to the OCFS volumes were shown as available, however, when trying
to umount and offline OCFS we see the behaviour outlined originally.
This is reproducible as follows -

1. Start with both paths to the OCFS volumes available
2. Shutdown the multipath daemon
3. Remove one path by disconnecting a cable to the switch
4. Restore the path by reconnecting cable
5. Start multipath daemon
6. Check both paths are available
7. Umount the OCFS file system (this returns immediately without the
usual few second delay)
8. Offline OCFS - error is received  

Under normal circumstances, i.e.: when the multipath daemon is
continuously available, losing a path, restoring a path, umount and
offline all work as expected.

Regards

Mark


Mark Schloss | Oracle DBA | Information Technology | x0013 

-----Original Message-----
From: Sunil Mushran [mailto:Sunil.Mushran at oracle.com] 
Sent: Tuesday, 22 July 2008 3:43 AM
To: Mark Schloss
Cc: ocfs2-users at oss.oracle.com
Subject: Re: [Ocfs2-users] OCFS processes active after a umount
[SEC=UNOFFICIAL]

That is strange.

Next time double check the mounts with:
$ cat /proc/mounts

The mount command prints the entries in /etc/mtab while
the /proc/mounts dumps the information from the kernel.
If those threads are there, it means the volume is still
mounted. Two in this case.

The entries in mtab are added by mount.ocfs2 and removed
by umount. There is a chance that mount.ocfs2 was unable
to add the entries in that file. Or, maybe one used the
-n option to force that behavior.

Which version/kernel is this?

Sunil

Mark Schloss wrote:
> Hello,
>  
> I have two OCFS file file systems mounted at /ocfs_1 and /ocfs_2. I 
> have unmounted both OCFS file systems and was trying to then offline 
> and unload OCFS. The offline command failed with -
>  
> # ./o2cb offline
> Stopping O2CB cluster ocfs2: Failed
> Unable to stop cluster as heartbeat region still active
> Looking at the processes on this box shows a number of OCFS processes 
> are still active -
>  
> ps -ef |grep ocf
> root      4704    23  0 Jul19 ?        00:00:00 [ocfs2_wq]
> root      4705    23  0 Jul19 ?        00:00:00 [ocfs2vote-0]
> root      4710    23  0 Jul19 ?        00:00:00 [ocfs2cmt-0]
> root      4730    23  0 Jul19 ?        00:00:00 [ocfs2vote-1]
> root      4735    23  0 Jul19 ?        00:00:00 [ocfs2cmt-1]
> root     10214  3485  0 18:12 pts/2    00:00:00 grep ocf
>  
> According to the FAQ, the ocfs2vote and ocfs2cmt processes should have

> gone at the umount.
>  
> Mount shows that there no OCFS file systems mounted -
>  
> # mount |grep -i ocfs
> ocfs2_dlmfs on /dlm type ocfs2_dlmfs (rw)
>  
> What can I do to offline/unload OCFS? (offline force fails with the 
> same message as above)
>  
> Regards
>  
> Mark Schloss

**********************************************************************
Please  note  that  your  email address  is known to  AUSTRAC  for the
purposes  of  communicating with you.  The information  transmitted in
this  e-mail is  for the  use of  the intended  recipient only and may
contain confidential and/or legally  privileged  material. If you have
received  this information  in error you must not disseminate, copy or
take  any  action on  it and we  request that you delete all copies of
this transmission together with attachments and notify the sender.

This footnote also confirms that this email message has been swept for
the presence of computer viruses.
**********************************************************************