Many thanks for managers for fast reply,
--- "Brusche, Johan" <johan.brusche_at_hp.com> wrote:
>
> Mahmod,
>
> The node in the second case is NOT the drd device
> server for the
> tape and possibly also NOT the cfs server for the
> filesystem being
> archived.
> Use drdmgr and cfsmgr to make the 2nd node the
> server for both.
>
> Johan.
>
> -----Oorspronkelijk bericht-----
> Van: tru64-unix-managers-owner_at_ornl.gov
> [mailto:tru64-unix-managers-owner_at_ornl.gov]Namens
> mahmod kokaje
> Verzonden: maandag 1 november 2004 14:30
> Aan: tru64-unix-managers_at_ornl.gov
> Onderwerp: tar problem
>
>
> Hi managers,
> on a cluster two es40 machines,memory channel
> interconnect,two HBA on eatch system,
> and msa6000 series tape library two drives,
> from one node I can do tar with buffer cahe like:
> #tar cvfBb /dev/ntape/tape1 10240 ..
> it fast for eatch tape drive,
> from other cluster system I cannt use the cache
> buffer
> like
> #tar cvfBb /dev/ntape/tape1 10240 ...
> becouse when issue this command the load on this
> node
> becom very heavy,
> any idea
> please advise.
>
>
>
>
> __________________________________
> Do you Yahoo!?
> Y! Messenger - Communicate in real time. Download
> now.
> http://messenger.yahoo.com
>
__________________________________
Do you Yahoo!?
Check out the new Yahoo! Front Page.
www.yahoo.com
Received on Tue Nov 02 2004 - 07:41:26 NZDT