Consistency and Standards – an IT Pro’s best bet in crisis

by admin on April 27, 2006 05:27pm

Seventeenth Century Philosopher and Author Voltaire wrote “I disapprove of what you say, but I will defend to the death your right to say it." This resonates very much with my own thoughts after going through the comments and feedback from my previous blog. I want to thank everyone for providing valuable feedback, regardless of whether I agree with it or not. The principle here is not to choose a side but to put a process in place that allows for an open and honest dialogue and I am exhilarated at the results of this endeavor.

Moving on to “Consistency and Standards” – the short theme for today’s blog. One of the questions put to me recently was to share something that I might have benefited from, during my past life in IT Operations. As I had mentioned previously, I have been involved with IT Operations since 1989 in some shape or form. From a student who worked at the Academic Computing Services in Syracuse University managing Macintosh and Sun Sparc clusters all the way to the past three years when I was managing 7x24 support of Class-A Production Services like AD, DNS, WINS etc. for MSN Operations, I can vouch for the fact that consistency in implementation standards have saved the day countless number of times. The point here is, no matter what platform, toolset, operating system or application you may choose, developing standards towards consistent implementation of these will always reap into rewards towards a lower “TCO” or Total Cost of Ownership in terms of supportability.

I remember a few years ago we were battling the spread of one of the malicious worms across the Internet. We were in the middle of taking inventory of what the configuration of our Production servers providing these mission-critical Class-A services, spread all across the world looked like. We all realized that by adhering to a common toolset, standard SKU’s for Hardware as well as for the OS versions helped us reduce the deployment cycle of a patch from what had seemed like days to a matter of hours. You may ask – “Hey – how does that matter” ? Well, imagine writing a different script for each type of configuration and multiply that by the hundreds of servers spread across the world in eight different datacenters. That’s quite a chore, especially when time is not on your side, you’re facing a crisis and you only have limited number of resources you can muster up for support. You’ll also need to track success and failure of applying the patch across datacenters and monitor where additional attention is needed.

So, what does that mean?  That complex environments require standards to work “for and with” IT Administrators. Admittedly,“one-off” or “out-of-standards” configs are very easy to do if we’re trying to please a group or mend fences with a specific customer. But in reality, we’re doing them (our customers) a dis-service but putting their environment in harm’s way and increasing their risks. Why - because supportability of their environment is ultimately our responsibility. So…the more colors we put on the fence, the more painters we will need and the longer it will take……