headers
sachin patel | 3 Jan 2010 03:45
Picon
Favicon

Re: nfs config - help needed


This is how we have setup.

1. xfs (mount fs here you might have ext3 file system)
2. exportfs (export your file system)
3. IP (virtual IP)

You start in this order and heartbeat will automatically stop in reverse order.

you might need to tune up your exportfs script which is located in /usr/lib/ocf/resource.d/heartbeat/

I have script called nfsserver

I made modification on following code.

####################################################
#
#   export_mount()
#
#         Get the status of the NFS mount and return whether it shows up in the etab or not
#
export_mount ()
{
        ocf_log info "--------> Running export_mount ${OCF_RESKEY_export_path} $
{OCF_RESKEY_export_parameters}"
        rc=0;
        if grep "^${OCF_RESKEY_export_path}\b" /var/lib/nfs/etab; then
                ocf_log info "${OCF_RESKEY_export_path} is already exported, removing"
                unexport_mount
                $rc=$?
(Continue reading)

Permalink | Reply | 11 comments |
headers
Hunny Bunny | 4 Jan 2010 04:11
Picon
Favicon

Distributed file system with DRBD configuration?

Hello cluster folkz,
I managed to install and configure DRBD on 3 node cluster. Node1 is a primary, node2 is a secondary and node3
is a stacked-on-top-of "node1_node2" backup.

Now I would like to use distributed FS either OCFS2 or GFS2 for the direct concurrent access of MySQL, SVN and
other shared data from all three nodes.

If I enable allow-two-primaries for the node1 and node2 can I still use node3 as a stacked-on-top-of
"node1_node2" backup, or it is off the table in allow-two-primaries configuration schema?
 
I could not find any DRBD configuration example which would allow to use three nodes simultaneously
something like "allow-three-primaries" without "stacking on top of" thing.

Could  somebody please enlighten me on this issue.

Many thanks in advance,

Alex 

      
_______________________________________________
Linux-HA mailing list
Linux-HA <at> lists.linux-ha.org
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
Permalink | Reply | 0 comments |
headers
Visu | 4 Jan 2010 05:44
Picon
Favicon

Re: nfs config - help needed

Where I need the xfs mount to be mounted (nodeA or nodeB)?  

Because this NFS mount to be mounted always on both the nodes(that is my
need).  When nodeA goes down, this xfs mount to be still available as well
as nfs server from nodeA to be shifted to nodeB automatically, when nodeA
goes down.

How I have to handle mount point points?

Which resource script I have to use for exportfs on Heartbeat?

-----Original Message-----
From: linux-ha-bounces <at> lists.linux-ha.org
[mailto:linux-ha-bounces <at> lists.linux-ha.org] On Behalf Of sachin patel
Sent: Sunday, January 03, 2010 6:46 AM
To: heartbeat-linux-ha
Subject: Re: [Linux-HA] nfs config - help needed

This is how we have setup.

1. xfs (mount fs here you might have ext3 file system)

2. exportfs (export your file system)

3. IP (virtual IP)

You start in this order and heartbeat will automatically stop in reverse
order.

you might need to tune up your exportfs script which is located in
(Continue reading)

Permalink | Reply | 4 comments |
headers
Andrew Beekhof | 4 Jan 2010 10:11
Favicon
Gravatar

Re: How to turn on file logging in corosync / openais

Probably best to ask about corosync on the openais list (instead of
the list of its competitor ;-)

On Fri, Dec 18, 2009 at 11:21 PM, Hunny Bunny <tmphb <at> yahoo.com> wrote:
> Hello folkz,
> I'm new to this list and to the HA cluster configuration / installation. Please have mercy for the newbie.
>
> I'm trying to enable corosync / openais logging into /var/log/corosync file
>
> in my corosync.conf / openais.conf I have this for the logging:
>
> <-------------- snipped ----------------->
>
> logging {
>     fileline:             off
>     to_syslog:        no
>     to_stderr:         no
>     to_logfile:         yes
>     logfile:              /var/log/corosync
>     syslog_facility: daemon
>     debug:             on
>     timestamp:       on
> }
>
> <-------------- snipped ----------------->
>
> However, it outputs logging info into /var/log/messages and /var/log/warn files but not in
/var/log/corosync file.
>
> Could somebody please enlighten me what is wrong with my logging configuration.
(Continue reading)

Permalink | Reply | 0 comments |
headers
Andrew Beekhof | 4 Jan 2010 10:22
Favicon
Gravatar

Re: [Linux - HA] Auto failover when httpd stop

On Wed, Dec 23, 2009 at 6:07 AM, Chandrakant Solanki
<solanki.chandrakant <at> gmail.com> wrote:
> ---------- Forwarded message ----------
> From: Chandrakant Solanki <solanki.chandrakant <at> gmail.com>
> Date: Wed, Dec 23, 2009 at 10:36 AM
> Subject: Re: Linux-HA Digest, Vol 73, Issue 28
> To: linux-ha <at> lists.linux-ha.org
>
>
>
>> Your primitive has no monitor operation defined. How should the cluster
>> recognize that the httpd is not running any more?
>>
>> --
>> Dr. Michael Schwartzkopff
>> MultiNET Services GmbH
>> Addresse: Bretonischer Ring 7; 85630 Grasbrunn; Germany
>> Tel: +49 - 89 - 45 69 11 0
>> Fax: +49 - 89 - 45 69 11 21
>> mob: +49 - 174 - 343 28 75
>>
>> mail: misch <at> multinet.de
>> web: www.multinet.de
>>
>> Sitz der Gesellschaft: 85630 Grasbrunn
>> Registergericht: Amtsgericht M?nchen HRB 114375
>> Gesch?ftsf?hrer: G?nter Jurgeneit, Hubert Martens
>>
>> ---
>>
(Continue reading)

Permalink | Reply | 0 comments |
headers
Andrew Beekhof | 4 Jan 2010 10:25
Favicon
Gravatar

Re: Node can not recover back to the cluster?

On Thu, Dec 24, 2009 at 5:50 AM, Jason Lee <listen1001 <at> 163.com> wrote:
> Now i find a violent way to let it back to the cluster.
> I stop the heartbeat on vm-3-2 and rm /var/lib/heartbeat/cmr/*
> and then start heartbeat again,
> it works.
>
> i think there should be a better way to do this.

There's no need to delete /var/lib/heartbeat/crm/*
Simply restarting heartbeat was what made the difference.

Perhaps file a bug for the heartbeat project.
_______________________________________________
Linux-HA mailing list
Linux-HA <at> lists.linux-ha.org
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems
Permalink | Reply | 1 comment |
headers
Lars Ellenberg | 4 Jan 2010 11:50
Favicon

Re: Question about linux-ha.org

On Sun, Jan 03, 2010 at 01:30:06PM -0700, Alan Robertson wrote:
> JabberUser wrote:
>> Hi Alan!
>>
>> My name is Richard, I work as administrator, in our company we are 
>> using heartbeat, now I see that HA is no longer maintained :(.

The opposite is the case!

You may also be interessted in the announcement of
 http://www.planet-ha.org/#LINBIT+announces+stewardship+for+Heartbeat+code+base

>> If You can pls let me know why I also wanna know if project will be  
>> taken by someone else.

You are probably referring to the "screaming" notice on
www.linux-ha.org, which states the following:

	This web page is no longer maintained.
	Information presented here exists only to avoid breaking historical links.
	For maintained information, see the Linux-HA Reference Documentation.

Clearly this is about the _web page_, not the _project_.
and it tells you also where to look next.

See also the announcement for this web site "upgrade":
 http://lists.linux-ha.org/pipermail/linux-ha-dev/2009-December/016917.html

Please do not hesitate to followup with further questions or suggestions
on the mailing list, or via PM.
(Continue reading)

Permalink | Reply | 0 comments |
headers
Dejan Muhamedagic | 4 Jan 2010 14:31

Re: nfs config - help needed

Hi,

On Sun, Jan 03, 2010 at 02:45:35AM +0000, sachin patel wrote:
> 
> This is how we have setup.
> 
> 1. xfs (mount fs here you might have ext3 file system)
> 2. exportfs (export your file system)
> 3. IP (virtual IP)
> 
> You start in this order and heartbeat will automatically stop in reverse order.
> 
> you might need to tune up your exportfs script which is located in /usr/lib/ocf/resource.d/heartbeat/

exportfs? Can't find it in the repository. Where did you get that
script? Or is it something you developed?

> I have script called nfsserver
> 
> I made modification on following code.

Which code do you refer to here? Again, can't find it in
repositories.

Thanks,

Dejan

> ####################################################
> #
(Continue reading)

Permalink | Reply | 5 comments |
headers
sachin patel | 4 Jan 2010 18:11
Picon
Favicon

Re: nfs config - help needed


I took nfsserver script and modify for our need.

> Date: Mon, 4 Jan 2010 14:31:21 +0100
> From: dejanmm <at> fastmail.fm
> To: linux-ha <at> lists.linux-ha.org
> Subject: Re: [Linux-HA] nfs config - help needed
> 
> Hi,
> 
> On Sun, Jan 03, 2010 at 02:45:35AM +0000, sachin patel wrote:
> > 
> > This is how we have setup.
> > 
> > 1. xfs (mount fs here you might have ext3 file system)
> > 2. exportfs (export your file system)
> > 3. IP (virtual IP)
> > 
> > You start in this order and heartbeat will automatically stop in reverse order.
> > 
> > you might need to tune up your exportfs script which is located in /usr/lib/ocf/resource.d/heartbeat/
> 
> exportfs? Can't find it in the repository. Where did you get that
> script? Or is it something you developed?
> 
> > I have script called nfsserver
> > 
> > I made modification on following code.
> 
> Which code do you refer to here? Again, can't find it in
(Continue reading)

Permalink | Reply | 4 comments |
headers
sachin patel | 4 Jan 2010 18:17
Picon
Favicon

Re: nfs config - help needed


You do not want to mount /dev/sdb5 on both system.
think if you have mounted this on both node and exported out from nodeA via nfs. 
People are actively using it and someone from system login to nodeB and by mistake writes in to that
filesystem or removes it. your FS will get corrupted. you can be in serious trouble.

if you read more heartbeat is design for this use.

In my view you mount /dev/sdb5 on  NodeA export it out from nodeA and when nodeA fails nodeB will
automatically mounts /dev/sdb5 and export it out. That is why you want to use virtual IP and on client when 
you mount this nfs filesystem use that virtual ip.

I hope I am making some sense here. exportfs is not a script. check for script name "nfsserver". we have
modify that script and have added that code.

> From: visu100 <at> eim.ae
> To: linux-ha <at> lists.linux-ha.org
> Date: Mon, 4 Jan 2010 08:44:29 +0400
> Subject: Re: [Linux-HA] nfs config - help needed
> 
> Where I need the xfs mount to be mounted (nodeA or nodeB)?  
> 
>  
> 
> Because this NFS mount to be mounted always on both the nodes(that is my
> need).  When nodeA goes down, this xfs mount to be still available as well
> as nfs server from nodeA to be shifted to nodeB automatically, when nodeA
> goes down.
> 
>
(Continue reading)

Permalink | Reply | 3 comments |

Gmane