Hi, folks. Hope you and yours are OK after the events of this week. I
just wrapped up a TruCluster 5.1/5.1a class at ZKO today and came up with
another question that stumped me. This one concerns niff and the clients
that use it in a TruCluster environment.
Situation: you have a cluster alias (default or additional) that is
assiged to the IP address of a NetRain, what is the failover sequence when
there is a problem with the interfaces? In other words, when niff posts
the EVM "down" event, do netrain and aliasd initiate their failover
simultaneously or serially?
Illustrative example. nr0 (consisting of ee0 and ee1) at 192.168.1.5 is
the alias IP address for alias clu_oracle. Someone pulls the plug on ee0's
cable when it is the active interace for nr0. Now, when the "down" event
is posted, does aliasd wait for netrain, or do both of them initiate their
failovers simultaneously?
Cheers,
Ed.
Edward J. Branley, edward_at_muffeletta.com, ed_at_softadv.com
ICQ 7800171
HomePage/PGP Key available at
http://www.seashell-software.com/ejb
Tru64 UNIX/TruCluster Server Instructor, Institute for Software Advancement
www.softadv.com, www.tru64stuff.com
Received on Fri Sep 14 2001 - 16:03:54 NZST