5 Questions To Ask Your Network Admin

0 Comments

Consulting in 2018 is an interesting gig.  Perhaps one day, twenty years ago, an expert could walk into a room expecting to be drilled with questions - after all, he or she was the expert, right?  Today, the power of Google and the availability of information makes my role as an 'expert' a different animal.  The tables are turned - I find myself asking the questions of very competent IT professionals, doing my best to provide value in this age of the Internet.

 

Siberia-Data-Center.jpg.jpg

Despite this, even the most well-educated IT admin can rarely keep up with the complexity and change rate of IT, which has never been higher. Whether you're a CEO, IT manager, or systems admin, here are five questions to ask of your network admin to ensure your organization is reducing risk while maximizing utilization and efficiency.  (Incidentally, if you're a network admin, you should probably be asking yourself these questions!)

 

How much time per week do you spend on network changes, versus how much time is spent on network planning?

The answer may surprise you.  A McKinsey study done in 2016 showed that 95% of network changes are done manually, and 75% of OPEX budgets go toward changes and troubleshooting.  As if that wasn't enough, 70% of network problems are due to human error!  I've spent time on this blog discussing the power of automation, as I believe there has never been a better time to get our network admins out of the weeds and focused on how to use technology to enable business outcomes.

If our data center was lost, how much data would we lose?

This is called a Recovery Point Objective (RPO), and the answer may surprise you.  Most applications are only backed up every 24 hours, and that doesn't exclude mission-critical apps!  Businesses today can't afford to lose an entire day of emails, sales transactions, and customer communication.  If you can only afford 10 minutes of lost data for an application, ensure your environment is tuned for this.

 

New Call-to-action

 

If our data center was lost, how long would we be down?

This is called a Recovery Time Objective (RTO).  At the risk of sounding like an alarmist, a bad RTO can put a company out of business faster than any new competitor.   Still, even if you're only down for a day, how much does that cost your organization - both in hard dollars and reputation?  This is an analysis every organization needs to perform, and the answer should determine the budget to improve the RTO.

** I often hear organizations tout a two-hour recovery process, which sounds amazing.  Unfortunately, this one requires a follow-up question:

How often do we test our DR plan?

When asking this question, be sure to look for more than a checkbox.  Spinning up non-critical apps in a DR site doesn't constitute a DR test.  The only way to know your true RTO is to perform a full test of every application, and this should be performed every 6-12 months with regularity.  After all, the network changes each week, if not every day, and a test from two years ago is highly unlikely to mean anything for your network as it's configured today.

 

Does our budget kill innovation?

Take the time to give your admin a whiteboard with the question "What would we do if our budget was doubled?"  Give the admin a chance to draw it out and explain it, and you may hear ideas you've never before heard.  Network admins are generally great at following budgets, but sometimes to a fault - they may never bother discussing beneficial new technologies that would break the current budget.  Allow your admin to dream a little, and you may find that a new technology pays for itself in ways the admin never realized it could.

I didn't like the answers I heard.  What do I do?

An old adage comes to mind - how does one eat an elephant?  One bite at a time.  Create a list of actionable items and arrange them by priorities.  Match this planning to current and future budget cycles, and make sure the executive team understands if current risk justifies an expanded budget.  Last, but not least, invite a so-called expert to the table to offer some insights.  I may not be Google, but I'd definitely enjoy whiteboarding all this with you.

 

Jeff Bio.jpg

Tags: Data Center

Written by Jeff Kish

Leave a Reply

    Related Post