Re: [Linux-HA] About OCFS2 and Pacemaker

2010-09-28 Thread Alain.Moulle
Hi, in fact, it was the udev rule for /dev/misc/ocfs2_control which was missing . Thanks. Alain > > but in this case, the script o2cb does not even start the daemon > > successfully : > > daemon err o2cb ERROR: Could not start /sbin/ocfs2_controld.pcmk > > > Is that where the daemon lives? >

Re: [Linux-HA] About OCFS2 and Pacemaker

2010-09-28 Thread Andrew Beekhof
On Mon, Sep 27, 2010 at 1:09 PM, Alain.Moulle wrote: > Hi > > I've all re-configured with dlm and o2cb as clone in Pacemaker > configuration, > but unfortunately, all works fine on one node, but as soon as the clone o2cb > starts on second node, both nodes crash . > > So I've got the last releases

Re: [Linux-HA] About OCFS2 and Pacemaker

2010-09-27 Thread Alain.Moulle
Sorry , the real error message is : daemon err o2cb ERROR: ocfs2_controld.pcmk did not come up Alain Hi I've all re-configured with dlm and o2cb as clone in Pacemaker configuration, but unfortunately, all works fine on one node, but as soon as the clone o2cb starts on second node, both nodes cr

Re: [Linux-HA] About OCFS2 and Pacemaker

2010-09-27 Thread Alain.Moulle
Hi I've all re-configured with dlm and o2cb as clone in Pacemaker configuration, but unfortunately, all works fine on one node, but as soon as the clone o2cb starts on second node, both nodes crash . So I've got the last releases of ocfs2 pcmk and the last module in kernel, all recompiled on e

Re: [Linux-HA] About OCFS2 and Pacemaker

2010-09-22 Thread Andrew Beekhof
On Wed, Sep 22, 2010 at 10:24 AM, Alain.Moulle wrote: > Hi again, > > I've have a look in the external/sdb device, it seems to be useful > only if we don't use a "by network" fencing solution , so as I'm > using fence_ipmilan, perhaps this shared device is not useful > in my case , right ? Correc

Re: [Linux-HA] About OCFS2 and Pacemaker

2010-09-22 Thread Alain.Moulle
Hi again, I've have a look in the external/sdb device, it seems to be useful only if we don't use a "by network" fencing solution , so as I'm using fence_ipmilan, perhaps this shared device is not useful in my case , right ? Regards ALain ___ Linux-HA m

Re: [Linux-HA] About OCFS2 and Pacemaker

2010-09-22 Thread Andrew Beekhof
On Wed, Sep 22, 2010 at 9:49 AM, Alain.Moulle wrote: > Hi Andrew, > > sorry to come again on this subject, I was about to switch to OCFS2 in > Pacemaker but : > > For configuration , I'm using the documentation "Oracle Cluster File > System 2", > I think it is always correct/valid ? > > It seems t

Re: [Linux-HA] About OCFS2 and Pacemaker

2010-09-22 Thread Alain.Moulle
Hi Andrew, sorry to come again on this subject, I was about to switch to OCFS2 in Pacemaker but : For configuration , I'm using the documentation "Oracle Cluster File System 2", I think it is always correct/valid ? It seems that o2cb/dlm need an external/sdb (with /dev/sdb2 partition) "for he

Re: [Linux-HA] About OCFS2 and Pacemaker

2010-09-21 Thread Alain.Moulle
Hi Andrew, ok I'm convinced now. I'll change the configuration. Thanks Alain > A bunch of filesystem guys and cluster guys got together and thought > through all the corners cases[1] that were possible and decided that > having a resource manager in charge of a cluster FS that was using an > intern

Re: [Linux-HA] About OCFS2 and Pacemaker

2010-09-21 Thread Andrew Beekhof
On Tue, Sep 21, 2010 at 12:48 PM, Alain.Moulle wrote: >> >> sorry but I don't fully understand : >> > >> > ?- I don't think there is any "fencing" functionnality in the OCFS2 >> > management, >> >> >> there is suicide, but that is unrelated to my point >> > Ok I didn't not know, but this could not

Re: [Linux-HA] About OCFS2 and Pacemaker

2010-09-21 Thread Alain.Moulle
> > sorry but I don't fully understand : > > > > ?- I don't think there is any "fencing" functionnality in the OCFS2 > > management, > > > there is suicide, but that is unrelated to my point > Ok I didn't not know, but this could not be disturbing for Pacemaker , could it ? It's quite as if

Re: [Linux-HA] About OCFS2 and Pacemaker

2010-09-21 Thread Andrew Beekhof
On Tue, Sep 21, 2010 at 11:13 AM, Alain.Moulle wrote: > Hi Andrew, > > sorry but I don't fully understand : > >  - I don't think there is any "fencing" functionnality in the OCFS2 > management, there is suicide, but that is unrelated to my point >   so for me the membership information remains o

Re: [Linux-HA] About OCFS2 and Pacemaker

2010-09-21 Thread Alain.Moulle
Hi Andrew, sorry but I don't fully understand : - I don't think there is any "fencing" functionnality in the OCFS2 management, so for me the membership information remains only at Pacemaker/corosync level. - I want all the FS OCFS2 always mounted on both sides , so from Pacemaker point o

Re: [Linux-HA] About OCFS2 and Pacemaker

2010-09-20 Thread Andrew Beekhof
On Mon, Sep 20, 2010 at 5:55 PM, Alain.Moulle wrote: > Hi > > I have a "philosophic" question about two nodes with FS under OCFS2 > and Pacemaker/corosync for the HA of both nodes. > > My choice was to let OCFS2 stack out of Pacemaker configuration, > so I let the services o2cb and ocfs2 started a

[Linux-HA] About OCFS2 and Pacemaker

2010-09-20 Thread Alain.Moulle
Hi I have a "philosophic" question about two nodes with FS under OCFS2 and Pacemaker/corosync for the HA of both nodes. My choice was to let OCFS2 stack out of Pacemaker configuration, so I let the services o2cb and ocfs2 started at boot time. And I only configure some FS of type OCFS2 as clone