Hi all,

I'm after some advice on mirroring volumes/pools between SANs.

More specifically, the customer has -
1) a 12 node stretched cluster (6 nodes in prod, 6 at a business
continuity site)
2) two IBM SANs (one at prod, one at BC site)
3) all components 2GB fibre connected.

Idea was to utilise NSS mirroring to mirror partitions at prod out to BC
site as this would allow for a 'nice' failover to BC site in case the
prod site was lost entirely (inc SAN).

Problem is that the partitions we're trying to software mirror are
between 544GB and 1800GB. During the mirroring process the server on
which the volumes are mounted are absolutely hammered (in terms of disk
I/O) and are causing a denial of service to the users.

Also, no matter what I do, I cannot seem to improve the throughput of the
mirroring process any better than 3MB/sec. At this speed, it will take 7
days to mirror the 1800GB volume and impact users for that period as well.

I'm after some advice from anyone who has had to deal with this situation
before. Ideas and comments welcome.