unixsyzadmin at gmail.com
Sun Feb 5 14:37:57 EST 2012
Thanks. The nodes don't have any VIP (Virtual IP) configured.
Each of the nodes run "gfs" and "fenced" service.
The type of storage is EMC, and if each node has to be able to access the
same disk simultaneously the SCSI reserves should be turned off. I am not
able to determine how the SCSI reserve's have been turned off and the
parameters on which the nodes are fenced.
# more /etc/redhat-release
Red Hat Enterprise Linux AS release 4 (Nahant Update 7)
# rpm -qa | egrep -i 'gfs|fence'
On Sat, Feb 4, 2012 at 10:12 PM, raj sourabh <rajsourabh1 at gmail.com> wrote:
> Node redundancy on redhat cluster depends on how cluster has been
> configured. VIP has to be configured for node level redundancy.
> It seems ,in your setup all the three nodes access same storage and GFS has
> been setup in the same. GFS allows concurent access to same disk from
> different nodes. However disk level redundancy has to be taken care from
> the storage side.
> Hope this helps.
> On Sat, Feb 4, 2012 at 11:54 PM, unix syzadmin <unixsyzadmin at gmail.com
> > Hi,
> > I have not worked on GFS before, but i am given additional
> > to support the existing GFS clusters and model the news ones after them.
> > While I am currently going through the redhat documentation; I wanted
> > explore other avenues for quick help on precisely documenting how the
> > existing GPFS clusters are setup.
> > The current setup has 3 Linux nodes running RHEL4 Update 7, all of them
> > allocated the same storage and access to the same GFS file-systems.
> > I am not sure if this is setup like a cluster and if can sustain a disk
> > a node failure.
> > Please point or suggest me in the right direction.
> > Thanks.
> > --
> > redhat-list mailing list
> > unsubscribe mailto:redhat-list-request at redhat.com?subject=unsubscribe
> > https://www.redhat.com/mailman/listinfo/redhat-list
> redhat-list mailing list
> unsubscribe mailto:redhat-list-request at redhat.com?subject=unsubscribe
More information about the redhat-list