Re: OPS and raw filesystems

Johnny Chan wrote:

> that's assuming that the machine with the failed instance is up for
> you to ftp or import from. there are a few alternatives to consider:

All disks are on Storage Arrays, and each OPS server node has access to all SAs.

By "import", I meant using the Veritas Volume Manager "import" command. Since this would only be done on the case of a node failure, the failing node would not have done a "deport", so we'd have to do a "forced import".

> o make sure the filesystem for your archive threads are on the shared
> SparcArray, not the individual node's local disk. If they are on
> a shared volume, you can mounted the failed instance's archive FS
> on to the surviving instance's node, if the failed instance's node is
> down (ie, not holding an exclusive filesysystem mount on the archive FS).
> If they are on the local disk of the failed machine, you are pretty much
> hosed unless you can reboot the machine or reconnect some disk cables.
 

> o set up a separate NFS server node whose job is to act as the fileserver
> for the PDB nodes' archive files. Thus, each node would NFS mount both
> archive thread FS. The caveat here is the network traffic you will introduce
> and the bandwidth issues of dealing with a very high redo switch rate.

No need to NFS mount them. We just "force import" all required filesystems.

-- 
Garry D. Robbins             |mailto:Garry.Robbins_at_Labatt.Com |Opinions
Senior Tech. Supp. Specialist|WWW:   http://www.labatt.com/   |expressed
Labatt Breweries of Canada   |Phone: (519) 667-7131           |are mine
London, Ont. CANADA N6A 4M3  |FAX:   (519) 667-7394           |alone.