[Linux-cluster] Freeze with cmirror and cLVM

[nagius]

Hi,

Here is my cluster.conf :

<cluster alias="dev_s2" config_version="3" name="dev_s2">
        <clusternodes>
                <clusternode name="xen0dev40" nodeid="1"
votes="1"></clusternode>
                <clusternode name="xen0dev41" nodeid="2"
votes="1"></clusternode>
                <clusternode name="xen0dev42" nodeid="3"
votes="1"></clusternode>
                <clusternode name="xen0dev43" nodeid="4"
votes="1"></clusternode>
                <clusternode name="xen0dev44" nodeid="5"
votes="1"></clusternode>
        </clusternodes>
</cluster>

Fencing has been disabled a boot time in /etc/sysconfig/cman :
 FENCE_JOIN="no"

Thanks,
Nicolas AGIUS


-----Message d'origine-----
De : linux-cluster-bounces at redhat.com
[mailto:linux-cluster-bounces at redhat.com] De la part de SATHYA - IT
Envoyé : mardi 20 décembre 2011 05:15
À : linux-cluster at redhat.com
Objet : Re: [Linux-cluster] Linux-cluster Digest, Vol 92, Issue 14

Hi Nicolos,

Can you post your cluster.conf file. Do you have the unfence option enabled
in your cluster configuration. At times, system hangs out without allowing
the access to files during this scenario.

Thanks

Sathya Narayanan V
Solution Architect	
M +91 9940680173 |T +91 44 42199500  | Service Desk +91 44 42199521
SERVICE - In PRECISION IT is a PASSION
----------------------------------------------------------------------------
-----------------------------
Precision Infomatic (M) Pvt Ltd
22, 1st Floor, Habibullah Road, T. Nagar, Chennai - 600 017. India.
www.precisionit.co.in

-----Original Message-----
Date: Mon, 19 Dec 2011 17:06:13 +0100
From: nagius at astek.fr
To: linux-cluster at redhat.com
Subject: [Linux-cluster] Freeze with cmirror and cLVM
Message-ID: <20111219170613.6wazswzvu8sw4cwc at webmail.astek.fr>
Content-Type: text/plain; charset="iso-8859-1"; Format="flowed";
	DelSp="Yes"

Hello all,

We've got a problem using a mirrored and clustered LV within a 5 nodes
cluster.

Everything is fine, except when a node leave the cluster, with cman_tool
leave, /etc/init/d/cman stop, electric poweroff or whatever.
In this case the storage just freeze on the others nodes, during 20 seconds,
until openais take into account the configuration change.

We can see this during the freeze in /var/log/messages :

?openais[3136]: [TOTEM] The token was lost in the OPERATIONAL state.
?[...]
?kernel: device-mapper: dm-log-clustered: [1lXq0Alc] Request timed
out: [DM_CLOG_MARK_REGION/405288] - retrying ?[...]
?openais[3136]: [TOTEM] entering OPERATIONAL state.

I'm running lastest 5.7, with :
?- kernel-xen-2.6.18-274.12.1,
?- cman-2.0.115-85,
?- lvm2-cluster-2.02.84-6
?- cmirror-1.1.39-10
?- kmod-cmirror-xen-0.1.22-3

Is this a normal behavior ?

Thanks for help
Nicolas AGIUS
?

--
Linux-cluster mailing list
Linux-cluster at redhat.com
https://www.redhat.com/mailman/listinfo/linux-cluster