Does anyone know if this is supported setup by Novell?
Thanks,

On Sun, 14 Dec 2008 14:55:40 +0000, Arun wrote:

> Does anyone know if this is supported setup by Novell? Thanks,

To translate:

GroupWise 7.0.3

Hosted on clustered OES2 - using Novell Cluster Services

The resource file system used will be Ext3.

Will it work, yes. On the Ext3 you will want to reduce the inodes down
to say 4k so that it matches NSS. On a stand alone server you would
disable atime - on a clustered setup this option is not available and
therefore the performance of the file system is compromised. With
OES2SP1 this is Incorporated within NSS and because of the direct access
from the host NSS does not interact with eDir which is the biggest slow
down - hence NSS is faster than Ext3 for hosting GroupWise.

HTH



--

This is exactly the setup I have built. Four-node OES2/SLES_SP1 cluster,
GW 7.0.1 on an EXT3 volume. After much research I concluded that EXT3
would be better than Novell's recommended reiserfs. It works
fine.....however I am having such trouble with the NCP Server object for
the volume that I may just change it to NSS, your comments here tipping
me in that direction.....


--
kmeuskens
------------------------------------------------------------------------
kmeuskens's Profile: http://forums.novell.com/member.php?userid=34199
View this thread: http://forums.novell.com/showthread.php?t=354211

We use ext3 file system on a shared disk (SAN) with a two node SLES 10
SP2 cluster running GroupWise 7.03 and it works very well.

If I were you I'd avoid using NCS as it is total junk. I use Heartbeat
(which is distributed with SLES 10) which works very well and is a lot
more robust and easier to manage than NCS which is very half-baked.

On Mon, 05 Jan 2009 23:35:54 +0000, Simon Tideswell wrote:

> We use ext3 file system on a shared disk (SAN) with a two node SLES 10
> SP2 cluster running GroupWise 7.03 and it works very well.
>
> If I were you I'd avoid using NCS as it is total junk. I use Heartbeat
> (which is distributed with SLES 10) which works very well and is a lot
> more robust and easier to manage than NCS which is very half-baked.

An interesting point of view - having setup, administered and upgraded
both Heartbeat and NCS there is no doubt in my mind what so ever - NCS
wins hands down... OCFS2 - automatically managed and resources are
available without having to edit a script or bounce a node - try that
with HA2 :-)

Each has it's own benefits...

T



--

Tim Heywood NSC SysOp;1710513 Wrote:
> On Mon, 05 Jan 2009 23:35:54 +0000, Simon Tideswell wrote:
>
> > We use ext3 file system on a shared disk (SAN) with a two node SLES
> 10
> > SP2 cluster running GroupWise 7.03 and it works very well.
> >
> > If I were you I'd avoid using NCS as it is total junk. I use
> Heartbeat
> > (which is distributed with SLES 10) which works very well and is a
> lot
> > more robust and easier to manage than NCS which is very half-baked.
>
> An interesting point of view - having setup, administered and upgraded
> both Heartbeat and NCS there is no doubt in my mind what so ever - NCS
> wins hands down... OCFS2 - automatically managed and resources are
> available without having to edit a script or bounce a node - try that
> with HA2 :-)
>
> Each has it's own benefits...
>
> T
>
>
>
> --

I prefer NCS on SLES10 over HB myself. Its easy to setup and works
great for the poison pill to the Linux server when a file system or
network connection outage occurs. Trying to Stonith a Linux server
under HB seems very messy.


--
ehanley
------------------------------------------------------------------------
ehanley's Profile: http://forums.novell.com/member.php?userid=2016
View this thread: http://forums.novell.com/showthread.php?t=354211