Comments (1)
  • Add a Comment
  • Edit
  • More Actions v
  • Quarantine this Entry

1 localhost commented Trackback

I have a simple HACMP question for one of the experts.<div>&nbsp;</div> We are migrating from Compaq Tru*Cluster to AIX+ GPFS + HACMP.<div>&nbsp;</div> The configuration is typically a two node cluster (p550) with fibre channel based shared disk array (DS4300).<div>&nbsp;</div> Typically we run applications on one node and upon failover the apps run on the second server. Failover times have to be very quick, under a minute or so. <div>&nbsp;</div> Is this possible with HACMP? We keep hearing conflicting reports that fast failover is only possible if we use raw devices and otherwise it can take 5 minutes. This is a bit of a surprise because on Tru*64 we could failover very fast while using the the cluster file system.<div>&nbsp;</div> Thanks very much for your answer.