On 11/02/07, Sean Craig - Snr Solutions Consultant - enigmatis <[EMAIL PROTECTED]> wrote: > > Hi Riyaz, > > The problem you have here is around one simple fact - without a cluster file > system, both nodes cannot mount the same partition (SAN volume) at the same > time. This would result in a corrupted file system. > > You have some options: > > 1 - use a clustered file system (look here > http://www.yolinux.com/TUTORIALS/LinuxClustersAndFileSystems.html, they don't > mention Veritas Cluster Volume Manager) > > (be careful here as you don't want a distributed file system which is > something totally different) > This can be usefull, but does not address the enabling/disabling of services) > > 2 - use automatic or manual failover procedures > > These include cluster software that will automatically mount/unmount volumes > according to process execution, or manually do it yourself. You can setup > mount points in the /etc/fstab with 'noauto' that will prevent your failover > node mounting the SAN volume at boot time. Then write (and properly test) > scripts to kill the primary node and enable the backup node. > > Regards > Sean > HI Riyaz,
Sorry Sean got here before I did! I guess he must have responded whilst I was typing (Sean you smart fellow) it however looks like we are singing from the same hymn sheet. Sean specialises in storage solutions so his advise on this topic will be golden! (sean you owe me a b33r now mate for the shameless plug of your l33t IT skillz) daN -- _____________________________________________________________ " They that can give up liberty to obtain a little temporary safety deserve neither liberty nor safety. Benjamin Franklin 1706 - 1790
