On 11/27/06, Vali Dragnuta <[EMAIL PROTECTED]> wrote:
On Mon, 2006-11-27 at 19:30 +0200, Tiberiu ATUDOREI wrote:
> Si tot eu mai adaug: single point of failure?
Daca e vorba sa fie storage comun probabil ca cel mai bine se scoate cu
un clusterFS care ii poate asigura atit redundanta necesara cit si
bandwidthul necesar.
Ex : http://www.clusterfs.com/ :
"Lustre features production-quality stability and failover, with zero
single points of failure. More than 100 teraflops worldwide are deployed
in production environments using Lustre to manage their data."
lustre e overkill pentru asa ceva, si mai mult are nevoie de un
master. OCFS sau GFS1-6.1/GFS2 (care are locking distrbuit, nu are
nevoie de master) sunt mult mai bune pentru cazul de 2-3 servere
pentru storage.
_______________________________________________
RLUG mailing list
[email protected]
http://lists.lug.ro/mailman/listinfo/rlug
