I'm looking into 2 node failover cluster and I'm bit confused here.

>AVS can currently only replicate data out of a cluster. When AVS is
>deployed in a cluster environment it asks for a shared storage location
>to store its configuration DB. Question is whether one cannot fool it,

Isn't the point of AVS to replicate data from a primary node to a 
secondary node, where these nodes together form failover pair of 
servers, that is, a single cluster?

>  That's right! That's the "problem" we need to fix to have a full
>non-shared solution. Like Steve said, AVS requires a shared-disc for
>failover between nodes, but that is not our goal with the "non-shared
>ZFS/NFS HA" solution. I did talk with Jim Dunhan about it, but did not

But isn't the point of your blog instructions at 
http://www.posix.brte.com.br/blog/?p=73 to circumvent this?

When I look at the AVS demos at 
http://opensolaris.org/os/project/avs/Demos/, I don't see any mention 
of Sun Cluster. What exactly is the role of Sun Cluster such that it 
is required and there couldn't be an alternative mechanism for 
managing HA services such as Linux-HA, http://www.linux-ha.org/, 
which presumably can run on x86 Solaris?
-- 

Maurice Volaski, mvolaski at aecom.yu.edu
Computing Support, Rose F. Kennedy Center
Albert Einstein College of Medicine of Yeshiva University

Reply via email to