[Pacemaker] ping_pong test fails

[Andrew Beekhof]

On Sat, Dec 10, 2011 at 12:56 PM, David Gilbert
<david.gilbert at august.com.au> wrote:
> Hi Uwe,
>
> Three things:
>
> 1. It's a known bug since Fedora 13 that's been carried over into
> RHEL/CentOS - see https://bugzilla.redhat.com/show_bug.cgi?id=632178.
>
> 2. Red Hat (and by extension CentOS) have deprecated dlm_controld.pcmk
> and gfs_controld.pcmk. The packages are not available in RHEL 6.1 or the
> CentOS CR repository.
>
> 3. The pdf version of "Clusters from Scratch" is out of sync with the
> online html version.

I believe I fixed that a couple of weeks ago.

> The html version takes into account the deprecation
> of dlm_controld.pcmk and gfs_controld.pcmk and the associated changes to
> the cluster messaging stack on RHEL-like distros (cman vs corosync).
>
> My suggestion - update your CentOS installs with the CentOS CR repo and
> follow the html version of "Clusters from Scratch".
>
> HTH
> David
>
>> Date: Fri, 09 Dec 2011 15:04:38 +0100
>> From: Andreas Kurz <andreas at hastexo.com>
>> To: The Pacemaker cluster resource manager
>>       <pacemaker at oss.clusterlabs.org>
>> Subject: Re: [Pacemaker] ping_pong test fails
>> Message-ID: <4EE21576.9 at hastexo.com>
>> Content-Type: text/plain; charset="iso-8859-1"
>>
>> Hello Uwe,
>>
>> On 12/07/2011 04:52 PM, Uwe Ritzschke wrote:
>> > Hi,
>> >
>> > Perhaps someone can help me? I've got a problem similar to the one
>> > posted by Raphael on 02/22/2010 [gfs2: ping_pong test fails]:
>> >
>> > When I'm running samba's ping_pong test on a 3-node pacemaker
> cluster
>> > with gfs2, it returns
>> >
>> > # ./ping_pong /srv/digedag/foo 3
>>
>> Don't know what's the impact, but ping_pong lock coherence testing
>> should have at least N-nodes+1 as parameter ... did you try:
>>
>> # ./ping_pong /srv/digedag/foo 4
>>
>>
>> Regards,
>> Andreas
>>
>> --
>> Need help with Pacemaker?
>> http://www.hastexo.com/now
>>
>> > lock at 1 failed! - Function not implemented
>> > unlock at 0 failed! - Function not implemented
>> > [..]
>> >
>> > Both dlm_controld.pcmk and gfs_controld.pcmk are running, mounting
>> seems
>> > to work correctly.
>> >
>> > Did I miss something crucial in the configuration (attached below)?
>> I've
>> > read the "Clusters from Scratch" guide over and over again, but I'm
> a
>> > bit lost.
>> >
>> > Any advice would be greatly appreciated!
>> >
>> > Thanks,
>> > Uwe
>> >
>> >
>> >
>> > ----Our setup----
>> > 3 nodes running
>> > centOS 6.0
>> > corosync 1.2.3
>> > pacemaker 1.1.2
>> > gfs-pcmk 3.0.12
>> > resource-agents 3.0.12
>> >
>> > Shared storage is a SAN
>> >
>> > ----pacemaker----
>> > primitive digedagFS ocf:heartbeat:Filesystem \
>> >     params device="/dev/mapper/digedag1p1" directory="/srv/digedag"
>> > fstype="gfs2" \
>> >     op start interval="0" timeout="60" \
>> >     op stop interval="0" timeout="60"
>> > primitive dlm ocf:pacemaker:controld \
>> >     op monitor interval="120" \
>> >     op start interval="0" timeout="90" \
>> >     op stop interval="0" timeout="100"
>> > primitive gfs-control ocf:pacemaker:controld \
>> >     params daemon="gfs_controld.pcmk" args="-g 0" \
>> >     op monitor interval="120" \
>> >     op start interval="0" timeout="90" \
>> >     op stop interval="0" timeout="100"
>> > clone digedagFSClone digedagFS
>> > clone dlm-clone dlm \
>> >     meta interleave="true"
>> > clone gfs-clone gfs-control \
>> >     meta interleave="true"
>> > colocation digedagFS-with-gfs-control inf: digedagFSClone gfs-clone
>> > colocation gfs-with-dlm inf: gfs-clone dlm-clone
>> > order start-digedagFS-after-gfs-control inf: gfs-clone
> digedagFSClone
>> > order start-gfs-after-dlm inf: dlm-clone gfs-clone
>> >
>> > ----/sys/log/messages----
>> > [...]
>> > Dec  7 16:04:14 localhost kernel: GFS2 (built Jun 27 2011 18:09:43)
>> > installed
>> > Dec  7 16:04:14 localhost kernel: GFS2: fsid=: Trying to join
> cluster
>> > "lock_dlm", "pcmk:smb"
>> > Dec  7 16:04:14 localhost kernel: dlm: Using TCP for communications
>> > Dec  7 16:04:14 localhost kernel: GFS2: fsid=pcmk:smb.0: Joined
>> cluster.
>> > Now mounting FS...
>> > Dec  7 16:04:14 localhost kernel: GFS2: fsid=pcmk:smb.0: jid=0,
>> already
>> > locked for use
>> > Dec  7 16:04:14 localhost kernel: GFS2: fsid=pcmk:smb.0: jid=0:
>> Looking
>> > at journal...
>> > Dec  7 16:04:14 localhost kernel: GFS2: fsid=pcmk:smb.0: jid=0: Done
>> > Dec  7 16:04:14 localhost kernel: GFS2: fsid=pcmk:smb.0: jid=1:
>> Trying
>> > to acquire journal lock...
>> > Dec  7 16:04:14 localhost kernel: GFS2: fsid=pcmk:smb.0: jid=1:
>> Looking
>> > at journal...
>> > Dec  7 16:04:14 localhost kernel: GFS2: fsid=pcmk:smb.0: jid=1: Done
>> > Dec  7 16:04:14 localhost kernel: GFS2: fsid=pcmk:smb.0: jid=2:
>> Trying
>> > to acquire journal lock...
>> > Dec  7 16:04:14 localhost kernel: GFS2: fsid=pcmk:smb.0: jid=2:
>> Looking
>> > at journal...
>> > Dec  7 16:04:14 localhost kernel: GFS2: fsid=pcmk:smb.0: jid=2: Done
>> >
>> > ----mount----
>> > # mount | grep digedag
>> > /dev/mapper/digedag1p1 on /srv/digedag type gfs2
>> > (rw,seclabel,relatime,hostdata=jid=0)
>> >
>> >
>> >
>> >
>> > _______________________________________________
>> > Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
>> > http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>> >
>> > Project Home: http://www.clusterlabs.org
>> > Getting started:
>> http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
>> > Bugs: http://bugs.clusterlabs.org
>>
>>
>>
>> -------------- next part --------------
>> A non-text attachment was scrubbed...
>> Name: signature.asc
>> Type: application/pgp-signature
>> Size: 286 bytes
>> Desc: OpenPGP digital signature
>> URL:
>>
> <http://oss.clusterlabs.org/pipermail/pacemaker/attachments/20111209/e1
>> 16afe0/attachment-0001.sig>
>
> _______________________________________________
> Pacemaker mailing list: Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker
>
> Project Home: http://www.clusterlabs.org
> Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://bugs.clusterlabs.org