- by Dan Perry, CBCP

The advent of Global Area Networks and equipment redundancy combined with the emergence of new technologies like Storage Area Networks (SAN’s) and Network Attached Storage (NAS) pose some very interesting IT contingency planning alternatives.
At AMD, contingency planning philosophy for key business critical systems entails installing redundant computer equipment, including SAN or NAS storage solutions, in at least two separate buildings. This type of setup allows a mirrored approach to be utilized providing full system redundancy in separate locations. It provides the ability to recover from loss of the primary data center and cause only momentary interruption to the using community. Because of reasonable fiber optic distance limitations, however, this separation will not, in the near future, be in excess of one mile.
In many situations where the user base is local, this is adequate because any regional disaster spanning an area that large will also shut down the using community of the systems involved and render the lack of system support moot.
This local area approach does however lose credence if the using community is disbursed around the globe. In that case, if afflicted by a regional disaster (i.e. earthquake), perfectly healthy, innocent environments in other lands around the world may be impacted, from an IT standpoint, to the same degree as ground zero of the disaster.
Obviously, this could inflict a painful, even deadly blow to the entire company.
I’m proposing a Paradigm Shift which, hopefully will help change our perception of IT contingency planning forever.
Today’s Wide Area Network capabilities are more than adequate for total user interface and functions such as local printing activity and virtually allow the user base to be thousands of miles remote from the physical computer installation.
At AMD, as I’m sure is true with many other global companies, most critical systems are distributed around the world, i.e. shipping systems are located in Asia Pacific, Europe and the United States: ERP environments are located in the United States and Europe: Desktop services systems are located in Asia Pacific, Europe and the United States.
Each of the specific regional examples have, or plan to have, redundant computer equipment located in separate buildings in the local area.
In many cases the secondary equipment is typically used as development and test environments which would be relinquished in the event a disaster should occur causing failure of the primary environment. The development systems would then be recruited into production support activity.
Picture the scenario of a major earthquake destroying two square miles where both northern California shipping support system data centers are located, but does not affect the shipping, receiving and warehouse activity being performed five miles away.
What’s to keep us from bringing that application up on our Singapore facilities secondary shipping support systems? The network is capable of handling the user interaction and invoice printing located in the northern California warehouse. Assuming Singapore does not have a coincidental disaster of their own. In that case California and possibly Singapore will have to move to the secondary shipping support system environments of facilities in Penang and Bangkok, and so on.
Obviously all like applications support environments must adhere to rigid standards allowing transfer of other sites to be easily implemented in any like environment. Adequate current data must be maintained in neutral locations to insure the ability of restoring necessary files and data bases.
To successfully achieve these results, a stipulation says that all like systems must be sized to handle the largest environment.
Isn’t this fun picturing the awesome recovery capability we would have with just a little common sense and contingency planning for the future.
It’s all a matter of thinking “out of the box” and expanding your vision to encompass contemporary systems and networking capabilities.
I will wager to say that, if not told, the user community might never be aware that their physical computer support had moved half way around the world.
Maybe all of you have already had this type of revelation and are well on the way of really taking advantage of the global attributes of today’s business. Maybe you are utilizing the distance between the various facilities in the company’s benefit instead of being overwhelmed with the complexities of communicating and synchronizing activities between locations in different geographic areas with diverse culture, language and time zones.

Maybe!!
Done properly there can be great contingency planning benefits gained by treating your company as if it were one again, much like the time when the whole company was housed within one metropolitan area.
With today’s communications ability all of this is possible and I believe should be the stated direction for the next decade.

Dan Perry, CBCP, has managed different computer Systems support organizations, (applications, systems and hardware) in excess of twenty years. He has been in management with AMD for approximately thirteen years and currently holds position of Sr. IT Staff member responsibility for IT Disaster Readiness worldwide.

«BACK to the Articles Index