Focus: Hurricane Sandy
system-recovery lessons learned. Very
few of our customers are all-physical or
all-virtual; they run hybrid or mixed environments. But their applications can have
complex interdependencies. This means
that managing change becomes vital. If
the recovery environment isn’t an accurate
reflection of the production environment,
the recovery will fail.
Avoid Last-Minute Alterations
to Your Hot Site
In a disaster situation, you want to
avoid negotiating last–minute changes
to your contracted hot site even if, over
time, you didn’t maintain production and
recovery environment changes. Making
changes only serves to delay your recovery even more. During Hurricane Sandy,
nearly one-third of our customers required
significant changes to their recovery
system’s specifications. These changes
included having to obtain higher-end servers, additional disk capacity, and different
tape technology and firewall configurations, among other things.
One common recovery pitfall was failing to consider all three layers of application tiers: database, middleware, and Web.
Some customers couldn’t recover all three
tiers. They perhaps found a way to get
the Web tier back but not the middleware
or database tier, or vice versa. Failing to
recover all three tiers leaves you stranded
just as effectively as failing to move your
data to a secondary site.
As for network connectivity, a remedy
developed. Those companies who
reviewed their core network design and
established potential failover routes were
able to avoid network congestion around
the storm-affected areas.
Internal Resource Planning
Proves Critical
The importance of concentrating
on internal resource planning can’t be
emphasized enough. This is especially
true with communicating to employees in
advance of an impending disaster about
taking personal measures to cope. These
issues include how to get to the data
center and what must be considered when
evacuating a family out of harm’s way,
among others.
