Mailing List Archive: Linux-HA: Users

Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)

Index | Next | Previous | View Threaded

alain.moulle at bull

Jun 12, 2012, 7:41 AM

Post #1 of 17 (1705 views)
Permalink

Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)

Hi

I tried to make a Pacemaker/cman stack working following the instructions
here :
(on Red-Hat 6.2)

So I stopped corosync and Pacemaker
I remove the corosync.conf
(and there were no /etc/corosync/service/pcmk file)

I've patched the cman.conf :
sed -i.sed "s/.*CMAN_QUORUM_TIMEOUT=.*/CMAN_QUORUM_TIMEOUT=0/g"
/etc/sysconfig/cman

I create the /etc/cluster/cluster.conf (and put it on both nodes chili2 &
chili3):
<?xml version="1.0"?>
<cluster config_version="1" name="chili2_chili3">
<logging debug="off"/>
<clusternodes>
<clusternode name="chili2" nodeid="1">
<fence>
<method name="pcmk-redirect">
<device name="pcmk" port="chili2"/>
</method>
</fence>
</clusternode>
<clusternode name="chili3" nodeid="2">
<fence>
<method name="pcmk-redirect">
<device name="pcmk" port="chili3"/>
</method>
</fence>
</clusternode>
</clusternodes>
<fencedevices>
<fencedevice name="pcmk" agent="fence_pcmk"/>
</fencedevices>
</cluster>

but when I start cman, all daemon start successfully,
Starting cluster:
Checking if cluster has been disabled at boot... [ OK ]
Checking Network Manager... [ OK ]
Global setup... [ OK ]
Loading kernel modules... [ OK ]
Mounting configfs... [ OK ]
Starting cman... [ OK ]
Waiting for quorum... [ OK ]
Starting fenced... [ OK ]
Starting dlm_controld... [ OK ]
Starting gfs_controld... [ OK ]
Unfencing self... [ OK ]
Joining fence domain... [ OK ]
and I got :
cman_tool nodes
Node Sts Inc Joined Name
1 M 16 2012-06-12 15:20:53 chili2
2 M 20 2012-06-12 15:30:27 chili3

but :
crm_mon -1
Connection to cluster failed: connection failed

I can't find what I have missed ...

Except if it is not expected to work on 6.2 ?

I have installed the following rpms :
openais-1.1.1-7.bl6.x86_64.rpm
openaislib-1.1.1-7.bl6.x86_64.rpm
modcluster-0.16.2-10.bl6.x86_64.rpm
ricci-0.16.2-35.bl6.x86_64.rpm
fence-virtd-checkpoint-0.2.1-8.bl6.x86_64.rpm
fence-virtd-multicast-0.2.1-8.bl6.x86_64.rpm
fence-virtd-0.2.1-8.bl6.x86_64.rpm
fence-virt-0.2.1-8.bl6.x86_64.rpm
(but not gfs2 and ocfs2 as I don't need them)

Thanks for help
Alain
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

andrew at beekhof

Jun 12, 2012, 6:13 PM

Post #2 of 17 (1692 views)
Permalink

Re: Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2) [In reply to]

On Wed, Jun 13, 2012 at 12:41 AM, <alain.moulle [at] bull> wrote:
> Hi
>
> I tried to make a Pacemaker/cman stack working following the instructions
> here :
> (on Red-Hat 6.2)
>
> So I stopped corosync and Pacemaker
> I remove the corosync.conf
> (and there were no /etc/corosync/service/pcmk file)
>
> I've patched the cman.conf :
> sed -i.sed "s/.*CMAN_QUORUM_TIMEOUT=.*/CMAN_QUORUM_TIMEOUT=0/g"
> /etc/sysconfig/cman
>
> I create the /etc/cluster/cluster.conf (and put it on both nodes chili2 &
> chili3):
> <?xml version="1.0"?>
> <cluster config_version="1" name="chili2_chili3">
> <logging debug="off"/>
> <clusternodes>
> <clusternode name="chili2" nodeid="1">
> <fence>
> <method name="pcmk-redirect">
> <device name="pcmk" port="chili2"/>
> </method>
> </fence>
> </clusternode>
> <clusternode name="chili3" nodeid="2">
> <fence>
> <method name="pcmk-redirect">
> <device name="pcmk" port="chili3"/>
> </method>
> </fence>
> </clusternode>
> </clusternodes>
> <fencedevices>
> <fencedevice name="pcmk" agent="fence_pcmk"/>
> </fencedevices>
> </cluster>
>
> but when I start cman, all daemon start successfully,
> Starting cluster:
> Checking if cluster has been disabled at boot... [ OK ]
> Checking Network Manager... [ OK ]
> Global setup... [ OK ]
> Loading kernel modules... [ OK ]
> Mounting configfs... [ OK ]
> Starting cman... [ OK ]
> Waiting for quorum... [ OK ]
> Starting fenced... [ OK ]
> Starting dlm_controld... [ OK ]
> Starting gfs_controld... [ OK ]
> Unfencing self... [ OK ]
> Joining fence domain... [ OK ]
> and I got :
> cman_tool nodes
> Node Sts Inc Joined Name
> 1 M 16 2012-06-12 15:20:53 chili2
> 2 M 20 2012-06-12 15:30:27 chili3
>
> but :
> crm_mon -1
> Connection to cluster failed: connection failed
>
> I can't find what I have missed ...

service pacemaker start :-)

> Except if it is not expected to work on 6.2 ?
>
> I have installed the following rpms :
> openais-1.1.1-7.bl6.x86_64.rpm
> openaislib-1.1.1-7.bl6.x86_64.rpm
> modcluster-0.16.2-10.bl6.x86_64.rpm
> ricci-0.16.2-35.bl6.x86_64.rpm
> fence-virtd-checkpoint-0.2.1-8.bl6.x86_64.rpm
> fence-virtd-multicast-0.2.1-8.bl6.x86_64.rpm
> fence-virtd-0.2.1-8.bl6.x86_64.rpm
> fence-virt-0.2.1-8.bl6.x86_64.rpm
> (but not gfs2 and ocfs2 as I don't need them)
>
> Thanks for help
> Alain
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

alain.moulle at bull

Jun 13, 2012, 12:09 AM

Post #3 of 17 (1692 views)
Permalink

Re: Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2) [In reply to]

Hi Andrew

Obviously ... now after your response ;-) Sorry but I thougth it was
started by cman as it was with service corosync start !
But I think the service pacemaker start is missing in the doc between
service cman start
and crm_mon -1 ;-)
http://www.clusterlabs.org/doc/en-US/Pacemaker/1.1/html/Clusters_from_Scratch/_bringing_the_cluster_online_with_cman.html

But ok it starts now, I 'll test this stack.
Thanks a lot Andrew.
Alain

De : Andrew Beekhof <andrew [at] beekhof>
A : General Linux-HA mailing list <linux-ha [at] lists>
Date : 13/06/2012 03:13
Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
Envoyé par : linux-ha-bounces [at] lists

On Wed, Jun 13, 2012 at 12:41 AM, <alain.moulle [at] bull> wrote:
> Hi
>
> I tried to make a Pacemaker/cman stack working following the
instructions
> here :
> (on Red-Hat 6.2)
>
> So I stopped corosync and Pacemaker
> I remove the corosync.conf
> (and there were no /etc/corosync/service/pcmk file)
>
> I've patched the cman.conf :
> sed -i.sed "s/.*CMAN_QUORUM_TIMEOUT=.*/CMAN_QUORUM_TIMEOUT=0/g"
> /etc/sysconfig/cman
>
> I create the /etc/cluster/cluster.conf (and put it on both nodes chili2
&
> chili3):
> <?xml version="1.0"?>
> <cluster config_version="1" name="chili2_chili3">
> <logging debug="off"/>
> <clusternodes>
> <clusternode name="chili2" nodeid="1">
> <fence>
> <method name="pcmk-redirect">
> <device name="pcmk" port="chili2"/>
> </method>
> </fence>
> </clusternode>
> <clusternode name="chili3" nodeid="2">
> <fence>
> <method name="pcmk-redirect">
> <device name="pcmk" port="chili3"/>
> </method>
> </fence>
> </clusternode>
> </clusternodes>
> <fencedevices>
> <fencedevice name="pcmk" agent="fence_pcmk"/>
> </fencedevices>
> </cluster>
>
> but when I start cman, all daemon start successfully,
> Starting cluster:
> Checking if cluster has been disabled at boot... [ OK ]
> Checking Network Manager... [ OK ]
> Global setup... [ OK ]
> Loading kernel modules... [ OK ]
> Mounting configfs... [ OK ]
> Starting cman... [ OK ]
> Waiting for quorum... [ OK ]
> Starting fenced... [ OK ]
> Starting dlm_controld... [ OK ]
> Starting gfs_controld... [ OK ]
> Unfencing self... [ OK ]
> Joining fence domain... [ OK ]
> and I got :
> cman_tool nodes
> Node Sts Inc Joined Name
> 1 M 16 2012-06-12 15:20:53 chili2
> 2 M 20 2012-06-12 15:30:27 chili3
>
> but :
> crm_mon -1
> Connection to cluster failed: connection failed
>
> I can't find what I have missed ...

service pacemaker start :-)

> Except if it is not expected to work on 6.2 ?
>
> I have installed the following rpms :
> openais-1.1.1-7.bl6.x86_64.rpm
> openaislib-1.1.1-7.bl6.x86_64.rpm
> modcluster-0.16.2-10.bl6.x86_64.rpm
> ricci-0.16.2-35.bl6.x86_64.rpm
> fence-virtd-checkpoint-0.2.1-8.bl6.x86_64.rpm
> fence-virtd-multicast-0.2.1-8.bl6.x86_64.rpm
> fence-virtd-0.2.1-8.bl6.x86_64.rpm
> fence-virt-0.2.1-8.bl6.x86_64.rpm
> (but not gfs2 and ocfs2 as I don't need them)
>
> Thanks for help
> Alain
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

andrew at beekhof

Jun 13, 2012, 12:22 AM

Post #4 of 17 (1685 views)
Permalink

Re: Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2) [In reply to]

On Wed, Jun 13, 2012 at 5:09 PM, <alain.moulle [at] bull> wrote:
> Hi Andrew
>
> Obviously ... now after your response ;-) Sorry but I thougth it was
> started by cman as it was with service corosync start !
> But I think the service pacemaker start is missing in the doc between
> service cman start

Good point. My bad :)

> and crm_mon -1 ;-)
> http://www.clusterlabs.org/doc/en-US/Pacemaker/1.1/html/Clusters_from_Scratch/_bringing_the_cluster_online_with_cman.html
>
> But ok it starts now, I 'll test this stack.
> Thanks a lot Andrew.
> Alain
>
>
>
> De : Andrew Beekhof <andrew [at] beekhof>
> A : General Linux-HA mailing list <linux-ha [at] lists>
> Date : 13/06/2012 03:13
> Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
> Envoyé par : linux-ha-bounces [at] lists
>
>
>
> On Wed, Jun 13, 2012 at 12:41 AM, <alain.moulle [at] bull> wrote:
>> Hi
>>
>> I tried to make a Pacemaker/cman stack working following the
> instructions
>> here :
>> (on Red-Hat 6.2)
>>
>> So I stopped corosync and Pacemaker
>> I remove the corosync.conf
>> (and there were no /etc/corosync/service/pcmk file)
>>
>> I've patched the cman.conf :
>> sed -i.sed "s/.*CMAN_QUORUM_TIMEOUT=.*/CMAN_QUORUM_TIMEOUT=0/g"
>> /etc/sysconfig/cman
>>
>> I create the /etc/cluster/cluster.conf (and put it on both nodes chili2
> &
>> chili3):
>> <?xml version="1.0"?>
>> <cluster config_version="1" name="chili2_chili3">
>> <logging debug="off"/>
>> <clusternodes>
>> <clusternode name="chili2" nodeid="1">
>> <fence>
>> <method name="pcmk-redirect">
>> <device name="pcmk" port="chili2"/>
>> </method>
>> </fence>
>> </clusternode>
>> <clusternode name="chili3" nodeid="2">
>> <fence>
>> <method name="pcmk-redirect">
>> <device name="pcmk" port="chili3"/>
>> </method>
>> </fence>
>> </clusternode>
>> </clusternodes>
>> <fencedevices>
>> <fencedevice name="pcmk" agent="fence_pcmk"/>
>> </fencedevices>
>> </cluster>
>>
>> but when I start cman, all daemon start successfully,
>> Starting cluster:
>> Checking if cluster has been disabled at boot... [ OK ]
>> Checking Network Manager... [ OK ]
>> Global setup... [ OK ]
>> Loading kernel modules... [ OK ]
>> Mounting configfs... [ OK ]
>> Starting cman... [ OK ]
>> Waiting for quorum... [ OK ]
>> Starting fenced... [ OK ]
>> Starting dlm_controld... [ OK ]
>> Starting gfs_controld... [ OK ]
>> Unfencing self... [ OK ]
>> Joining fence domain... [ OK ]
>> and I got :
>> cman_tool nodes
>> Node Sts Inc Joined Name
>> 1 M 16 2012-06-12 15:20:53 chili2
>> 2 M 20 2012-06-12 15:30:27 chili3
>>
>> but :
>> crm_mon -1
>> Connection to cluster failed: connection failed
>>
>> I can't find what I have missed ...
>
> service pacemaker start :-)
>
>> Except if it is not expected to work on 6.2 ?
>>
>> I have installed the following rpms :
>> openais-1.1.1-7.bl6.x86_64.rpm
>> openaislib-1.1.1-7.bl6.x86_64.rpm
>> modcluster-0.16.2-10.bl6.x86_64.rpm
>> ricci-0.16.2-35.bl6.x86_64.rpm
>> fence-virtd-checkpoint-0.2.1-8.bl6.x86_64.rpm
>> fence-virtd-multicast-0.2.1-8.bl6.x86_64.rpm
>> fence-virtd-0.2.1-8.bl6.x86_64.rpm
>> fence-virt-0.2.1-8.bl6.x86_64.rpm
>> (but not gfs2 and ocfs2 as I don't need them)
>>
>> Thanks for help
>> Alain
>> _______________________________________________
>> Linux-HA mailing list
>> Linux-HA [at] lists
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

alain.moulle at bull

Jun 13, 2012, 12:23 AM

Post #5 of 17 (1684 views)
Permalink

Re: Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2) [In reply to]

Hi again,

by the way : using cman with cluster.conf, which network is used for
heartbeat ?

the one which is linked to the 'uname -n' name ?

and we can't use two rings as with corosync ?

except if we use bond IF ?

Alain

De : Andrew Beekhof <andrew [at] beekhof>
A : General Linux-HA mailing list <linux-ha [at] lists>
Date : 13/06/2012 03:13
Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
Envoyé par : linux-ha-bounces [at] lists

On Wed, Jun 13, 2012 at 12:41 AM, <alain.moulle [at] bull> wrote:
> Hi
>
> I tried to make a Pacemaker/cman stack working following the
instructions
> here :
> (on Red-Hat 6.2)
>
> So I stopped corosync and Pacemaker
> I remove the corosync.conf
> (and there were no /etc/corosync/service/pcmk file)
>
> I've patched the cman.conf :
> sed -i.sed "s/.*CMAN_QUORUM_TIMEOUT=.*/CMAN_QUORUM_TIMEOUT=0/g"
> /etc/sysconfig/cman
>
> I create the /etc/cluster/cluster.conf (and put it on both nodes chili2
&
> chili3):
> <?xml version="1.0"?>
> <cluster config_version="1" name="chili2_chili3">
> <logging debug="off"/>
> <clusternodes>
> <clusternode name="chili2" nodeid="1">
> <fence>
> <method name="pcmk-redirect">
> <device name="pcmk" port="chili2"/>
> </method>
> </fence>
> </clusternode>
> <clusternode name="chili3" nodeid="2">
> <fence>
> <method name="pcmk-redirect">
> <device name="pcmk" port="chili3"/>
> </method>
> </fence>
> </clusternode>
> </clusternodes>
> <fencedevices>
> <fencedevice name="pcmk" agent="fence_pcmk"/>
> </fencedevices>
> </cluster>
>
> but when I start cman, all daemon start successfully,
> Starting cluster:
> Checking if cluster has been disabled at boot... [ OK ]
> Checking Network Manager... [ OK ]
> Global setup... [ OK ]
> Loading kernel modules... [ OK ]
> Mounting configfs... [ OK ]
> Starting cman... [ OK ]
> Waiting for quorum... [ OK ]
> Starting fenced... [ OK ]
> Starting dlm_controld... [ OK ]
> Starting gfs_controld... [ OK ]
> Unfencing self... [ OK ]
> Joining fence domain... [ OK ]
> and I got :
> cman_tool nodes
> Node Sts Inc Joined Name
> 1 M 16 2012-06-12 15:20:53 chili2
> 2 M 20 2012-06-12 15:30:27 chili3
>
> but :
> crm_mon -1
> Connection to cluster failed: connection failed
>
> I can't find what I have missed ...

service pacemaker start :-)

> Except if it is not expected to work on 6.2 ?
>
> I have installed the following rpms :
> openais-1.1.1-7.bl6.x86_64.rpm
> openaislib-1.1.1-7.bl6.x86_64.rpm
> modcluster-0.16.2-10.bl6.x86_64.rpm
> ricci-0.16.2-35.bl6.x86_64.rpm
> fence-virtd-checkpoint-0.2.1-8.bl6.x86_64.rpm
> fence-virtd-multicast-0.2.1-8.bl6.x86_64.rpm
> fence-virtd-0.2.1-8.bl6.x86_64.rpm
> fence-virt-0.2.1-8.bl6.x86_64.rpm
> (but not gfs2 and ocfs2 as I don't need them)
>
> Thanks for help
> Alain
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

andrew at beekhof

Jun 14, 2012, 9:35 PM

Post #6 of 17 (1673 views)
Permalink

Re: Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2) [In reply to]

On Wed, Jun 13, 2012 at 5:23 PM, <alain.moulle [at] bull> wrote:
> Hi again,
>
> by the way : using cman with cluster.conf, which network is used for
> heartbeat ?

That I don't know.
Try the red hat documentation?

>
> the one which is linked to the 'uname -n' name ?
>
> and we can't use two rings as with corosync ?

I believe its possible, I don't know the details though.

>
> except if we use bond IF ?
>
> Alain
>
>
>
> De : Andrew Beekhof <andrew [at] beekhof>
> A : General Linux-HA mailing list <linux-ha [at] lists>
> Date : 13/06/2012 03:13
> Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
> Envoyé par : linux-ha-bounces [at] lists
>
>
>
> On Wed, Jun 13, 2012 at 12:41 AM, <alain.moulle [at] bull> wrote:
>> Hi
>>
>> I tried to make a Pacemaker/cman stack working following the
> instructions
>> here :
>> (on Red-Hat 6.2)
>>
>> So I stopped corosync and Pacemaker
>> I remove the corosync.conf
>> (and there were no /etc/corosync/service/pcmk file)
>>
>> I've patched the cman.conf :
>> sed -i.sed "s/.*CMAN_QUORUM_TIMEOUT=.*/CMAN_QUORUM_TIMEOUT=0/g"
>> /etc/sysconfig/cman
>>
>> I create the /etc/cluster/cluster.conf (and put it on both nodes chili2
> &
>> chili3):
>> <?xml version="1.0"?>
>> <cluster config_version="1" name="chili2_chili3">
>> <logging debug="off"/>
>> <clusternodes>
>> <clusternode name="chili2" nodeid="1">
>> <fence>
>> <method name="pcmk-redirect">
>> <device name="pcmk" port="chili2"/>
>> </method>
>> </fence>
>> </clusternode>
>> <clusternode name="chili3" nodeid="2">
>> <fence>
>> <method name="pcmk-redirect">
>> <device name="pcmk" port="chili3"/>
>> </method>
>> </fence>
>> </clusternode>
>> </clusternodes>
>> <fencedevices>
>> <fencedevice name="pcmk" agent="fence_pcmk"/>
>> </fencedevices>
>> </cluster>
>>
>> but when I start cman, all daemon start successfully,
>> Starting cluster:
>> Checking if cluster has been disabled at boot... [ OK ]
>> Checking Network Manager... [ OK ]
>> Global setup... [ OK ]
>> Loading kernel modules... [ OK ]
>> Mounting configfs... [ OK ]
>> Starting cman... [ OK ]
>> Waiting for quorum... [ OK ]
>> Starting fenced... [ OK ]
>> Starting dlm_controld... [ OK ]
>> Starting gfs_controld... [ OK ]
>> Unfencing self... [ OK ]
>> Joining fence domain... [ OK ]
>> and I got :
>> cman_tool nodes
>> Node Sts Inc Joined Name
>> 1 M 16 2012-06-12 15:20:53 chili2
>> 2 M 20 2012-06-12 15:30:27 chili3
>>
>> but :
>> crm_mon -1
>> Connection to cluster failed: connection failed
>>
>> I can't find what I have missed ...
>
> service pacemaker start :-)
>
>> Except if it is not expected to work on 6.2 ?
>>
>> I have installed the following rpms :
>> openais-1.1.1-7.bl6.x86_64.rpm
>> openaislib-1.1.1-7.bl6.x86_64.rpm
>> modcluster-0.16.2-10.bl6.x86_64.rpm
>> ricci-0.16.2-35.bl6.x86_64.rpm
>> fence-virtd-checkpoint-0.2.1-8.bl6.x86_64.rpm
>> fence-virtd-multicast-0.2.1-8.bl6.x86_64.rpm
>> fence-virtd-0.2.1-8.bl6.x86_64.rpm
>> fence-virt-0.2.1-8.bl6.x86_64.rpm
>> (but not gfs2 and ocfs2 as I don't need them)
>>
>> Thanks for help
>> Alain
>> _______________________________________________
>> Linux-HA mailing list
>> Linux-HA [at] lists
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

alain.moulle at bull

Jun 15, 2012, 5:06 AM

Post #7 of 17 (1672 views)
Permalink

Re: Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2) [In reply to]

Hi Andrew

you recall me in an old thread here that effectively cman was not involved

in option 4 : corosync + cpg + quorumd + mcp
whereas it is involved in option 3 : corosync + cpg + cman + mcp
but is seems that corosync is also used in both options .

I tried to configure option 3 as you've seen in my other email two days
ago, and we
only have a mini cluster.conf file , and no more corosync.conf (and it
works once
I start Pacemaker after cman ;-) )

My question is now :
when the option 4 will be available, we will come back to the
corosync.conf file ?
as same as with option 2 and no more cluster.conf ?

And to be completely clear on why my question :
the temporary option 3 forces us to use a mini cluster.conf, and therefore
only one heartbeat network (or two but with bonding).
But if in the future we configure option 4, and come back to
corosync.conf, we will be able to have again two networks rings in
the corosync.conf, and so ... that sounds be much better for me.
Excpet if quorumd is working with a mini cluster.conf like cman ?

Thanks for these precisions.
Regards
Alain
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

andrew at beekhof

Jun 16, 2012, 3:25 AM

Post #8 of 17 (1666 views)
Permalink

Re: Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2) [In reply to]

On Fri, Jun 15, 2012 at 10:06 PM, <alain.moulle [at] bull> wrote:
> Hi Andrew
>
> you recall me in an old thread here that effectively cman was not involved
>
> in option 4 : corosync + cpg + quorumd + mcp
> whereas it is involved in option 3 : corosync + cpg + cman + mcp
> but is seems that corosync is also used in both options .

cman is just a corosync plugin. think of "cman" being an alias for
"corosync + cman plugin"

>
> I tried to configure option 3 as you've seen in my other email two days
> ago, and we
> only have a mini cluster.conf file , and no more corosync.conf (and it
> works once
> I start Pacemaker after cman ;-) )
>
> My question is now :
> when the option 4 will be available, we will come back to the
> corosync.conf file ?

yes

> as same as with option 2 and no more cluster.conf ?

right

> And to be completely clear on why my question :
> the temporary option 3 forces us to use a mini cluster.conf, and therefore
> only one heartbeat network (or two but with bonding).

I'm pretty sure you can have redundant rings with cluster.conf, I just
don't know the details.

> But if in the future we configure option 4, and come back to
> corosync.conf, we will be able to have again two networks rings in
> the corosync.conf, and so ... that sounds be much better for me.
> Excpet if quorumd is working with a mini cluster.conf like cman ?

No. Just corosync.conf
You can get a preview of how option 4 works here:
https://www.dropbox.com/s/zd1mi6u1m7ac5t9/Pacemaker-1.1-Clusters_from_Scratch-en-US.pdf

I need to finish it off and push to clusterlabs...

>
> Thanks for these precisions.
> Regards
> Alain
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

alain.moulle at bull

Jun 18, 2012, 1:11 AM

Post #9 of 17 (1673 views)
Permalink

Re: Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2) [In reply to]

Hi Andrew

Thanks a lot, this is now much clearer for me.

And yes it is possible to have redundant heartbeat with the option 3 and
cluster.conf, but only thanks to bonding functionnality :

man cman
Multi-home configuration
It is quite common to use multiple ethernet adapters for
cluster nodes, so they will tolerate the failure
of one link. A common way to do this is to use ethernet
bonding. Alternatively you can get corosync to
run in redundant ring mode by specifying an â€™altnameâ€™ for
the node. This is an alternative name by which
the node is known, that resolves to another IP address
used on the other ethernet adapter(s). You can
optionally specify a different port and/or multicast address
for each altname in use. Up to 9 altnames
(10 interfaces in total) can be used.

But my choice will be to keep Pacemker corosync plugin from option 2, even
on RH 6.3 , and to switch
directly to option 4 when it is available on RH 6.3 (therefore to skip the
option 3) that's the very best choice
for me, as option 4 is coming back with a corosync.conf.

Thanks a lot
Alain

De : Andrew Beekhof <andrew [at] beekhof>
A : General Linux-HA mailing list <linux-ha [at] lists>
Date : 16/06/2012 12:25
Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
EnvoyÃ© par : linux-ha-bounces [at] lists

On Fri, Jun 15, 2012 at 10:06 PM, <alain.moulle [at] bull> wrote:
> Hi Andrew
>
> you recall me in an old thread here that effectively cman was not
involved
>
> in option 4 : corosync + cpg + quorumd + mcp
> whereas it is involved in option 3 : corosync + cpg + cman + mcp
> but is seems that corosync is also used in both options .

cman is just a corosync plugin. think of "cman" being an alias for
"corosync + cman plugin"

>
> I tried to configure option 3 as you've seen in my other email two days
> ago, and we
> only have a mini cluster.conf file , and no more corosync.conf (and it
> works once
> I start Pacemaker after cman ;-) )
>
> My question is now :
> when the option 4 will be available, we will come back to the
> corosync.conf file ?

yes

> as same as with option 2 and no more cluster.conf ?

right

> And to be completely clear on why my question :
> the temporary option 3 forces us to use a mini cluster.conf, and
therefore
> only one heartbeat network (or two but with bonding).

I'm pretty sure you can have redundant rings with cluster.conf, I just
don't know the details.

> But if in the future we configure option 4, and come back to
> corosync.conf, we will be able to have again two networks rings in
> the corosync.conf, and so ... that sounds be much better for me.
> Excpet if quorumd is working with a mini cluster.conf like cman ?

No. Just corosync.conf
You can get a preview of how option 4 works here:

https://www.dropbox.com/s/zd1mi6u1m7ac5t9/Pacemaker-1.1-Clusters_from_Scratch-en-US.pdf

I need to finish it off and push to clusterlabs...

>
> Thanks for these precisions.
> Regards
> Alain
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

alain.moulle at bull

Jun 18, 2012, 6:04 AM

Post #10 of 17 (1660 views)
Permalink

Re: Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2) [In reply to]

Hi again,

could you tell me the package which install the "pacemaker plugin v1" and
which
is the name of the binary or binaries or src ?

Thanks a lot
Alain

De : Andrew Beekhof <andrew [at] beekhof>
A : General Linux-HA mailing list <linux-ha [at] lists>
Date : 16/06/2012 12:25
Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
Envoyé par : linux-ha-bounces [at] lists

On Fri, Jun 15, 2012 at 10:06 PM, <alain.moulle [at] bull> wrote:
> Hi Andrew
>
> you recall me in an old thread here that effectively cman was not
involved
>
> in option 4 : corosync + cpg + quorumd + mcp
> whereas it is involved in option 3 : corosync + cpg + cman + mcp
> but is seems that corosync is also used in both options .

cman is just a corosync plugin. think of "cman" being an alias for
"corosync + cman plugin"

>
> I tried to configure option 3 as you've seen in my other email two days
> ago, and we
> only have a mini cluster.conf file , and no more corosync.conf (and it
> works once
> I start Pacemaker after cman ;-) )
>
> My question is now :
> when the option 4 will be available, we will come back to the
> corosync.conf file ?

yes

> as same as with option 2 and no more cluster.conf ?

right

> And to be completely clear on why my question :
> the temporary option 3 forces us to use a mini cluster.conf, and
therefore
> only one heartbeat network (or two but with bonding).

I'm pretty sure you can have redundant rings with cluster.conf, I just
don't know the details.

> But if in the future we configure option 4, and come back to
> corosync.conf, we will be able to have again two networks rings in
> the corosync.conf, and so ... that sounds be much better for me.
> Excpet if quorumd is working with a mini cluster.conf like cman ?

No. Just corosync.conf
You can get a preview of how option 4 works here:

https://www.dropbox.com/s/zd1mi6u1m7ac5t9/Pacemaker-1.1-Clusters_from_Scratch-en-US.pdf

I need to finish it off and push to clusterlabs...

>
> Thanks for these precisions.
> Regards
> Alain
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

andreas at hastexo

Jun 18, 2012, 8:43 AM

Post #11 of 17 (1653 views)
Permalink

Re: Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2) [In reply to]

On 06/18/2012 03:04 PM, alain.moulle [at] bull wrote:
> Hi again,
>
> could you tell me the package which install the "pacemaker plugin v1" and
> which
> is the name of the binary or binaries or src ?

You only need to add:

service {
# Load the Pacemaker Cluster Resource Manager
ver: 1
name: pacemaker
}

to your corosync.conf ... or create a file with this content in
/etc/corosync/service.d/.

Once you started Corosync you need to start the pacemaker init script,
that starts the MCP ... and stop that services in reverse order.

The init script is part of the pacemaker package on RHEL 6.x

Regards,
Andreas

--
Need help with Pacemaker?
http://www.hastexo.com/now

>
> Thanks a lot
> Alain
>
>
>
> De : Andrew Beekhof <andrew [at] beekhof>
> A : General Linux-HA mailing list <linux-ha [at] lists>
> Date : 16/06/2012 12:25
> Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
> Envoyé par : linux-ha-bounces [at] lists
>
>
>
> On Fri, Jun 15, 2012 at 10:06 PM, <alain.moulle [at] bull> wrote:
>> Hi Andrew
>>
>> you recall me in an old thread here that effectively cman was not
> involved
>>
>> in option 4 : corosync + cpg + quorumd + mcp
>> whereas it is involved in option 3 : corosync + cpg + cman + mcp
>> but is seems that corosync is also used in both options .
>
> cman is just a corosync plugin. think of "cman" being an alias for
> "corosync + cman plugin"
>
>>
>> I tried to configure option 3 as you've seen in my other email two days
>> ago, and we
>> only have a mini cluster.conf file , and no more corosync.conf (and it
>> works once
>> I start Pacemaker after cman ;-) )
>>
>> My question is now :
>> when the option 4 will be available, we will come back to the
>> corosync.conf file ?
>
> yes
>
>> as same as with option 2 and no more cluster.conf ?
>
> right
>
>> And to be completely clear on why my question :
>> the temporary option 3 forces us to use a mini cluster.conf, and
> therefore
>> only one heartbeat network (or two but with bonding).
>
> I'm pretty sure you can have redundant rings with cluster.conf, I just
> don't know the details.
>
>> But if in the future we configure option 4, and come back to
>> corosync.conf, we will be able to have again two networks rings in
>> the corosync.conf, and so ... that sounds be much better for me.
>> Excpet if quorumd is working with a mini cluster.conf like cman ?
>
> No. Just corosync.conf
> You can get a preview of how option 4 works here:
>
> https://www.dropbox.com/s/zd1mi6u1m7ac5t9/Pacemaker-1.1-Clusters_from_Scratch-en-US.pdf
>
>
> I need to finish it off and push to clusterlabs...
>
>>
>> Thanks for these precisions.
>> Regards
>> Alain
>> _______________________________________________
>> Linux-HA mailing list
>> Linux-HA [at] lists
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>

Attachments:

signature.asc (0.22 KB)

andrew at beekhof

Jun 18, 2012, 2:38 PM

Post #12 of 17 (1649 views)
Permalink

Re: Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2) [In reply to]

On Mon, Jun 18, 2012 at 11:04 PM, <alain.moulle [at] bull> wrote:
> Hi again,
>
> could you tell me the package which install the "pacemaker plugin v1" and
> which
> is the name of the binary or binaries or src ?

its the pacemaker source rpm. it doesn't ship by default iirc

>
> Thanks a lot
> Alain
>
>
>
> De : Andrew Beekhof <andrew [at] beekhof>
> A : General Linux-HA mailing list <linux-ha [at] lists>
> Date : 16/06/2012 12:25
> Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
> Envoyé par : linux-ha-bounces [at] lists
>
>
>
> On Fri, Jun 15, 2012 at 10:06 PM, <alain.moulle [at] bull> wrote:
>> Hi Andrew
>>
>> you recall me in an old thread here that effectively cman was not
> involved
>>
>> in option 4 : corosync + cpg + quorumd + mcp
>> whereas it is involved in option 3 : corosync + cpg + cman + mcp
>> but is seems that corosync is also used in both options .
>
> cman is just a corosync plugin. think of "cman" being an alias for
> "corosync + cman plugin"
>
>>
>> I tried to configure option 3 as you've seen in my other email two days
>> ago, and we
>> only have a mini cluster.conf file , and no more corosync.conf (and it
>> works once
>> I start Pacemaker after cman ;-) )
>>
>> My question is now :
>> when the option 4 will be available, we will come back to the
>> corosync.conf file ?
>
> yes
>
>> as same as with option 2 and no more cluster.conf ?
>
> right
>
>> And to be completely clear on why my question :
>> the temporary option 3 forces us to use a mini cluster.conf, and
> therefore
>> only one heartbeat network (or two but with bonding).
>
> I'm pretty sure you can have redundant rings with cluster.conf, I just
> don't know the details.
>
>> But if in the future we configure option 4, and come back to
>> corosync.conf, we will be able to have again two networks rings in
>> the corosync.conf, and so ... that sounds be much better for me.
>> Excpet if quorumd is working with a mini cluster.conf like cman ?
>
> No. Just corosync.conf
> You can get a preview of how option 4 works here:
>
> https://www.dropbox.com/s/zd1mi6u1m7ac5t9/Pacemaker-1.1-Clusters_from_Scratch-en-US.pdf
>
>
> I need to finish it off and push to clusterlabs...
>
>>
>> Thanks for these precisions.
>> Regards
>> Alain
>> _______________________________________________
>> Linux-HA mailing list
>> Linux-HA [at] lists
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

alain.moulle at bull

Jun 19, 2012, 12:22 AM

Post #13 of 17 (1635 views)
Permalink

Re: Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2) [In reply to]

Hi

Sorry but I don't know what iirc means , I suppose iirc in this context
stands for "plugin".
If so, how can I check for sure that the plugin is or is not in the
pacemaker package ?
(it is to check the pacemaker package delivered with the new RH 6.3)

Thanks
Alain

De : Andrew Beekhof <andrew [at] beekhof>
A : General Linux-HA mailing list <linux-ha [at] lists>
Date : 18/06/2012 23:38
Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
Envoyé par : linux-ha-bounces [at] lists

On Mon, Jun 18, 2012 at 11:04 PM, <alain.moulle [at] bull> wrote:
> Hi again,
>
> could you tell me the package which install the "pacemaker plugin v1"
and
> which
> is the name of the binary or binaries or src ?

its the pacemaker source rpm. it doesn't ship by default iirc

>
> Thanks a lot
> Alain
>
>
>
> De : Andrew Beekhof <andrew [at] beekhof>
> A : General Linux-HA mailing list <linux-ha [at] lists>
> Date : 16/06/2012 12:25
> Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
> Envoyé par : linux-ha-bounces [at] lists
>
>
>
> On Fri, Jun 15, 2012 at 10:06 PM, <alain.moulle [at] bull> wrote:
>> Hi Andrew
>>
>> you recall me in an old thread here that effectively cman was not
> involved
>>
>> in option 4 : corosync + cpg + quorumd + mcp
>> whereas it is involved in option 3 : corosync + cpg + cman + mcp
>> but is seems that corosync is also used in both options .
>
> cman is just a corosync plugin. think of "cman" being an alias for
> "corosync + cman plugin"
>
>>
>> I tried to configure option 3 as you've seen in my other email two days
>> ago, and we
>> only have a mini cluster.conf file , and no more corosync.conf (and it
>> works once
>> I start Pacemaker after cman ;-) )
>>
>> My question is now :
>> when the option 4 will be available, we will come back to the
>> corosync.conf file ?
>
> yes
>
>> as same as with option 2 and no more cluster.conf ?
>
> right
>
>> And to be completely clear on why my question :
>> the temporary option 3 forces us to use a mini cluster.conf, and
> therefore
>> only one heartbeat network (or two but with bonding).
>
> I'm pretty sure you can have redundant rings with cluster.conf, I just
> don't know the details.
>
>> But if in the future we configure option 4, and come back to
>> corosync.conf, we will be able to have again two networks rings in
>> the corosync.conf, and so ... that sounds be much better for me.
>> Excpet if quorumd is working with a mini cluster.conf like cman ?
>
> No. Just corosync.conf
> You can get a preview of how option 4 works here:
>
>
https://www.dropbox.com/s/zd1mi6u1m7ac5t9/Pacemaker-1.1-Clusters_from_Scratch-en-US.pdf

>
>
> I need to finish it off and push to clusterlabs...
>
>>
>> Thanks for these precisions.
>> Regards
>> Alain
>> _______________________________________________
>> Linux-HA mailing list
>> Linux-HA [at] lists
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

andreas at hastexo

Jun 19, 2012, 1:26 AM

Post #14 of 17 (1649 views)
Permalink

Re: Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2) [In reply to]

On 06/19/2012 09:22 AM, alain.moulle [at] bull wrote:
> Hi
>
> Sorry but I don't know what iirc means , I suppose iirc in this context
> stands for "plugin".
> If so, how can I check for sure that the plugin is or is not in the
> pacemaker package ?
> (it is to check the pacemaker package delivered with the new RH 6.3)

Check the pacemaker package if it contains the "/etc/init.d/pacemaker"
init script, that start the MCP ... if yes, then adopt your
corosync.conf as I wrote in my previous mail.

Regards,
Andreas

--
Need help with Pacemaker?
http://www.hastexo.com/now

>
> Thanks
> Alain
>
>
>
> De : Andrew Beekhof <andrew [at] beekhof>
> A : General Linux-HA mailing list <linux-ha [at] lists>
> Date : 18/06/2012 23:38
> Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
> Envoyé par : linux-ha-bounces [at] lists
>
>
>
> On Mon, Jun 18, 2012 at 11:04 PM, <alain.moulle [at] bull> wrote:
>> Hi again,
>>
>> could you tell me the package which install the "pacemaker plugin v1"
> and
>> which
>> is the name of the binary or binaries or src ?
>
> its the pacemaker source rpm. it doesn't ship by default iirc
>
>>
>> Thanks a lot
>> Alain
>>
>>
>>
>> De : Andrew Beekhof <andrew [at] beekhof>
>> A : General Linux-HA mailing list <linux-ha [at] lists>
>> Date : 16/06/2012 12:25
>> Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
>> Envoyé par : linux-ha-bounces [at] lists
>>
>>
>>
>> On Fri, Jun 15, 2012 at 10:06 PM, <alain.moulle [at] bull> wrote:
>>> Hi Andrew
>>>
>>> you recall me in an old thread here that effectively cman was not
>> involved
>>>
>>> in option 4 : corosync + cpg + quorumd + mcp
>>> whereas it is involved in option 3 : corosync + cpg + cman + mcp
>>> but is seems that corosync is also used in both options .
>>
>> cman is just a corosync plugin. think of "cman" being an alias for
>> "corosync + cman plugin"
>>
>>>
>>> I tried to configure option 3 as you've seen in my other email two days
>>> ago, and we
>>> only have a mini cluster.conf file , and no more corosync.conf (and it
>>> works once
>>> I start Pacemaker after cman ;-) )
>>>
>>> My question is now :
>>> when the option 4 will be available, we will come back to the
>>> corosync.conf file ?
>>
>> yes
>>
>>> as same as with option 2 and no more cluster.conf ?
>>
>> right
>>
>>> And to be completely clear on why my question :
>>> the temporary option 3 forces us to use a mini cluster.conf, and
>> therefore
>>> only one heartbeat network (or two but with bonding).
>>
>> I'm pretty sure you can have redundant rings with cluster.conf, I just
>> don't know the details.
>>
>>> But if in the future we configure option 4, and come back to
>>> corosync.conf, we will be able to have again two networks rings in
>>> the corosync.conf, and so ... that sounds be much better for me.
>>> Excpet if quorumd is working with a mini cluster.conf like cman ?
>>
>> No. Just corosync.conf
>> You can get a preview of how option 4 works here:
>>
>>
> https://www.dropbox.com/s/zd1mi6u1m7ac5t9/Pacemaker-1.1-Clusters_from_Scratch-en-US.pdf
>
>>
>>
>> I need to finish it off and push to clusterlabs...
>>
>>>
>>> Thanks for these precisions.
>>> Regards
>>> Alain
>>> _______________________________________________
>>> Linux-HA mailing list
>>> Linux-HA [at] lists
>>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>>> See also: http://linux-ha.org/ReportingProblems
>> _______________________________________________
>> Linux-HA mailing list
>> Linux-HA [at] lists
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
>>
>> _______________________________________________
>> Linux-HA mailing list
>> Linux-HA [at] lists
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>

Attachments:

signature.asc (0.22 KB)

andrew at beekhof

Jun 19, 2012, 2:34 AM

Post #15 of 17 (1632 views)
Permalink

Re: Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2) [In reply to]

On Tue, Jun 19, 2012 at 5:22 PM, <alain.moulle [at] bull> wrote:
> Hi
>
> Sorry but I don't know what iirc means , I suppose iirc in this context
> stands for "plugin".

iirc == if i recall correctly :-)

> If so, how can I check for sure that the plugin is or is not in the
> pacemaker package ?
> (it is to check the pacemaker package delivered with the new RH 6.3)

'rpm -ql pacemaker' will give you the file list. look for something
in the corosync directories.

fyi though, even if it is present now, we're planning to remove it in
the next point release or so.
the policy is to not ship things we're not supporting.

>
> Thanks
> Alain
>
>
>
> De : Andrew Beekhof <andrew [at] beekhof>
> A : General Linux-HA mailing list <linux-ha [at] lists>
> Date : 18/06/2012 23:38
> Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
> Envoyé par : linux-ha-bounces [at] lists
>
>
>
> On Mon, Jun 18, 2012 at 11:04 PM, <alain.moulle [at] bull> wrote:
>> Hi again,
>>
>> could you tell me the package which install the "pacemaker plugin v1"
> and
>> which
>> is the name of the binary or binaries or src ?
>
> its the pacemaker source rpm. it doesn't ship by default iirc
>
>>
>> Thanks a lot
>> Alain
>>
>>
>>
>> De : Andrew Beekhof <andrew [at] beekhof>
>> A : General Linux-HA mailing list <linux-ha [at] lists>
>> Date : 16/06/2012 12:25
>> Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
>> Envoyé par : linux-ha-bounces [at] lists
>>
>>
>>
>> On Fri, Jun 15, 2012 at 10:06 PM, <alain.moulle [at] bull> wrote:
>>> Hi Andrew
>>>
>>> you recall me in an old thread here that effectively cman was not
>> involved
>>>
>>> in option 4 : corosync + cpg + quorumd + mcp
>>> whereas it is involved in option 3 : corosync + cpg + cman + mcp
>>> but is seems that corosync is also used in both options .
>>
>> cman is just a corosync plugin. think of "cman" being an alias for
>> "corosync + cman plugin"
>>
>>>
>>> I tried to configure option 3 as you've seen in my other email two days
>>> ago, and we
>>> only have a mini cluster.conf file , and no more corosync.conf (and it
>>> works once
>>> I start Pacemaker after cman ;-) )
>>>
>>> My question is now :
>>> when the option 4 will be available, we will come back to the
>>> corosync.conf file ?
>>
>> yes
>>
>>> as same as with option 2 and no more cluster.conf ?
>>
>> right
>>
>>> And to be completely clear on why my question :
>>> the temporary option 3 forces us to use a mini cluster.conf, and
>> therefore
>>> only one heartbeat network (or two but with bonding).
>>
>> I'm pretty sure you can have redundant rings with cluster.conf, I just
>> don't know the details.
>>
>>> But if in the future we configure option 4, and come back to
>>> corosync.conf, we will be able to have again two networks rings in
>>> the corosync.conf, and so ... that sounds be much better for me.
>>> Excpet if quorumd is working with a mini cluster.conf like cman ?
>>
>> No. Just corosync.conf
>> You can get a preview of how option 4 works here:
>>
>>
> https://www.dropbox.com/s/zd1mi6u1m7ac5t9/Pacemaker-1.1-Clusters_from_Scratch-en-US.pdf
>
>>
>>
>> I need to finish it off and push to clusterlabs...
>>
>>>
>>> Thanks for these precisions.
>>> Regards
>>> Alain
>>> _______________________________________________
>>> Linux-HA mailing list
>>> Linux-HA [at] lists
>>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>>> See also: http://linux-ha.org/ReportingProblems
>> _______________________________________________
>> Linux-HA mailing list
>> Linux-HA [at] lists
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
>>
>> _______________________________________________
>> Linux-HA mailing list
>> Linux-HA [at] lists
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

alain.moulle at bull

Jun 19, 2012, 2:43 AM

Post #16 of 17 (1638 views)
Permalink

Re: Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2) [In reply to]

> iirc == if i recall correctly :-)

sorry but I'm french, so ... I did not understand such achronymous ... but
now I'll know ;-)

> 'rpm -ql pacemaker' will give you the file list. look for something
> in the corosync directories.
yes I know all those linux commands, but even on the Pacemaker release I
currently use
and which works with the corosync.conf , so with the plugin v1, I can't
find anything
related to corosync in pacemaker rpm :
rpm -ql pacemaker | grep -i coro
returns nothing at all

> fyi though, even if it is present now, we're planning to remove it in
> the next point release or so.
> the policy is to not ship things we're not supporting.
Yes I know, but if it is always delivered on 6.3, that will be sufficient
for me until I switch from stack option 2 to stack option 4

Alain

De : Andrew Beekhof <andrew [at] beekhof>
A : General Linux-HA mailing list <linux-ha [at] lists>
Date : 19/06/2012 11:34
Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
Envoyé par : linux-ha-bounces [at] lists

> If so, how can I check for sure that the plugin is or is not in the
> pacemaker package ?
> (it is to check the pacemaker package delivered with the new RH 6.3)

'rpm -ql pacemaker' will give you the file list. look for something
in the corosync directories.

fyi though, even if it is present now, we're planning to remove it in
the next point release or so.
the policy is to not ship things we're not supporting.

>
> Thanks
> Alain
>
>
>
> De : Andrew Beekhof <andrew [at] beekhof>
> A : General Linux-HA mailing list <linux-ha [at] lists>
> Date : 18/06/2012 23:38
> Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
> Envoyé par : linux-ha-bounces [at] lists
>
>
>
> On Mon, Jun 18, 2012 at 11:04 PM, <alain.moulle [at] bull> wrote:
>> Hi again,
>>
>> could you tell me the package which install the "pacemaker plugin v1"
> and
>> which
>> is the name of the binary or binaries or src ?
>
> its the pacemaker source rpm. it doesn't ship by default iirc
>
>>
>> Thanks a lot
>> Alain
>>
>>
>>
>> De : Andrew Beekhof <andrew [at] beekhof>
>> A : General Linux-HA mailing list <linux-ha [at] lists>
>> Date : 16/06/2012 12:25
>> Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH
6.2)
>> Envoyé par : linux-ha-bounces [at] lists
>>
>>
>>
>> On Fri, Jun 15, 2012 at 10:06 PM, <alain.moulle [at] bull> wrote:
>>> Hi Andrew
>>>
>>> you recall me in an old thread here that effectively cman was not
>> involved
>>>
>>> in option 4 : corosync + cpg + quorumd + mcp
>>> whereas it is involved in option 3 : corosync + cpg + cman + mcp
>>> but is seems that corosync is also used in both options .
>>
>> cman is just a corosync plugin. think of "cman" being an alias for
>> "corosync + cman plugin"
>>
>>>
>>> I tried to configure option 3 as you've seen in my other email two
days
>>> ago, and we
>>> only have a mini cluster.conf file , and no more corosync.conf (and it
>>> works once
>>> I start Pacemaker after cman ;-) )
>>>
>>> My question is now :
>>> when the option 4 will be available, we will come back to the
>>> corosync.conf file ?
>>
>> yes
>>
>>> as same as with option 2 and no more cluster.conf ?
>>
>> right
>>
>>> And to be completely clear on why my question :
>>> the temporary option 3 forces us to use a mini cluster.conf, and
>> therefore
>>> only one heartbeat network (or two but with bonding).
>>
>> I'm pretty sure you can have redundant rings with cluster.conf, I just
>> don't know the details.
>>
>>> But if in the future we configure option 4, and come back to
>>> corosync.conf, we will be able to have again two networks rings in
>>> the corosync.conf, and so ... that sounds be much better for me.
>>> Excpet if quorumd is working with a mini cluster.conf like cman ?
>>
>> No. Just corosync.conf
>> You can get a preview of how option 4 works here:
>>
>>
>
https://www.dropbox.com/s/zd1mi6u1m7ac5t9/Pacemaker-1.1-Clusters_from_Scratch-en-US.pdf

>
>>
>>
>> I need to finish it off and push to clusterlabs...
>>
>>>
>>> Thanks for these precisions.
>>> Regards
>>> Alain
>>> _______________________________________________
>>> Linux-HA mailing list
>>> Linux-HA [at] lists
>>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>>> See also: http://linux-ha.org/ReportingProblems
>> _______________________________________________
>> Linux-HA mailing list
>> Linux-HA [at] lists
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
>>
>> _______________________________________________
>> Linux-HA mailing list
>> Linux-HA [at] lists
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

andrew at beekhof

Jun 19, 2012, 3:02 AM

Post #17 of 17 (1644 views)
Permalink

Re: Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2) [In reply to]

On Tue, Jun 19, 2012 at 7:43 PM, <alain.moulle [at] bull> wrote:
>> iirc == if i recall correctly :-)
>
> sorry but I'm french, so ... I did not understand such achronymous ... but
> now I'll know ;-)
>
>> 'rpm -ql pacemaker' will give you the file list. look for something
>> in the corosync directories.
> yes I know all those linux commands, but even on the Pacemaker release I
> currently use
> and which works with the corosync.conf , so with the plugin v1, I can't
> find anything
> related to corosync in pacemaker rpm :
> rpm -ql pacemaker | grep -i coro
> returns nothing at all

(Its been a long time since I've used the plugin, so I had to look it up)

You want:
%{_libexecdir}/lcrso/pacemaker.lcrso

which whould be in the main 'pacemaker' package if anywhere.

>
>> fyi though, even if it is present now, we're planning to remove it in
>> the next point release or so.
>> the policy is to not ship things we're not supporting.
> Yes I know, but if it is always delivered on 6.3, that will be sufficient
> for me until I switch from stack option 2 to stack option 4
>
> Alain
>
>
>
>
> De : Andrew Beekhof <andrew [at] beekhof>
> A : General Linux-HA mailing list <linux-ha [at] lists>
> Date : 19/06/2012 11:34
> Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
> Envoyé par : linux-ha-bounces [at] lists
>
>
>
>
>
>> If so, how can I check for sure that the plugin is or is not in the
>> pacemaker package ?
>> (it is to check the pacemaker package delivered with the new RH 6.3)
>
> 'rpm -ql pacemaker' will give you the file list. look for something
> in the corosync directories.
>
> fyi though, even if it is present now, we're planning to remove it in
> the next point release or so.
> the policy is to not ship things we're not supporting.
>
>>
>> Thanks
>> Alain
>>
>>
>>
>> De : Andrew Beekhof <andrew [at] beekhof>
>> A : General Linux-HA mailing list <linux-ha [at] lists>
>> Date : 18/06/2012 23:38
>> Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH 6.2)
>> Envoyé par : linux-ha-bounces [at] lists
>>
>>
>>
>> On Mon, Jun 18, 2012 at 11:04 PM, <alain.moulle [at] bull> wrote:
>>> Hi again,
>>>
>>> could you tell me the package which install the "pacemaker plugin v1"
>> and
>>> which
>>> is the name of the binary or binaries or src ?
>>
>> its the pacemaker source rpm. it doesn't ship by default iirc
>>
>>>
>>> Thanks a lot
>>> Alain
>>>
>>>
>>>
>>> De : Andrew Beekhof <andrew [at] beekhof>
>>> A : General Linux-HA mailing list <linux-ha [at] lists>
>>> Date : 16/06/2012 12:25
>>> Objet : Re: [Linux-HA] Pacemaker/corosync ==> Pacemaker/cman (on RH
> 6.2)
>>> Envoyé par : linux-ha-bounces [at] lists
>>>
>>>
>>>
>>> On Fri, Jun 15, 2012 at 10:06 PM, <alain.moulle [at] bull> wrote:
>>>> Hi Andrew
>>>>
>>>> you recall me in an old thread here that effectively cman was not
>>> involved
>>>>
>>>> in option 4 : corosync + cpg + quorumd + mcp
>>>> whereas it is involved in option 3 : corosync + cpg + cman + mcp
>>>> but is seems that corosync is also used in both options .
>>>
>>> cman is just a corosync plugin. think of "cman" being an alias for
>>> "corosync + cman plugin"
>>>
>>>>
>>>> I tried to configure option 3 as you've seen in my other email two
> days
>>>> ago, and we
>>>> only have a mini cluster.conf file , and no more corosync.conf (and it
>>>> works once
>>>> I start Pacemaker after cman ;-) )
>>>>
>>>> My question is now :
>>>> when the option 4 will be available, we will come back to the
>>>> corosync.conf file ?
>>>
>>> yes
>>>
>>>> as same as with option 2 and no more cluster.conf ?
>>>
>>> right
>>>
>>>> And to be completely clear on why my question :
>>>> the temporary option 3 forces us to use a mini cluster.conf, and
>>> therefore
>>>> only one heartbeat network (or two but with bonding).
>>>
>>> I'm pretty sure you can have redundant rings with cluster.conf, I just
>>> don't know the details.
>>>
>>>> But if in the future we configure option 4, and come back to
>>>> corosync.conf, we will be able to have again two networks rings in
>>>> the corosync.conf, and so ... that sounds be much better for me.
>>>> Excpet if quorumd is working with a mini cluster.conf like cman ?
>>>
>>> No. Just corosync.conf
>>> You can get a preview of how option 4 works here:
>>>
>>>
>>
> https://www.dropbox.com/s/zd1mi6u1m7ac5t9/Pacemaker-1.1-Clusters_from_Scratch-en-US.pdf
>
>>
>>>
>>>
>>> I need to finish it off and push to clusterlabs...
>>>
>>>>
>>>> Thanks for these precisions.
>>>> Regards
>>>> Alain
>>>> _______________________________________________
>>>> Linux-HA mailing list
>>>> Linux-HA [at] lists
>>>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>>>> See also: http://linux-ha.org/ReportingProblems
>>> _______________________________________________
>>> Linux-HA mailing list
>>> Linux-HA [at] lists
>>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>>> See also: http://linux-ha.org/ReportingProblems
>>>
>>> _______________________________________________
>>> Linux-HA mailing list
>>> Linux-HA [at] lists
>>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>>> See also: http://linux-ha.org/ReportingProblems
>> _______________________________________________
>> Linux-HA mailing list
>> Linux-HA [at] lists
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
>>
>> _______________________________________________
>> Linux-HA mailing list
>> Linux-HA [at] lists
>> http://lists.linux-ha.org/mailman/listinfo/linux-ha
>> See also: http://linux-ha.org/ReportingProblems
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
>
> _______________________________________________
> Linux-HA mailing list
> Linux-HA [at] lists
> http://lists.linux-ha.org/mailman/listinfo/linux-ha
> See also: http://linux-ha.org/ReportingProblems
_______________________________________________
Linux-HA mailing list
Linux-HA [at] lists
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Index | Next | Previous | View Threaded

Interested in having your list archived? Contact Gossamer Threads