Ten 2020 Resolutions for a Data Center Manager

'Tis the season when we all—well most—set resolutions for the new year. I did some thinking over the holidays and came up with a list that data center manager’s might be able to relate to. So, if you haven’t yet developed your list for 2020, need to add a couple more, or just want to compare, here are 10 suggestions.  Hopefully, you’ll walk away with a nugget or two you can use.  

1. Document 2019 Wins!   We get so busy looking ahead, we forget to take a bit of time to document our wins from the previous year.  Hopefully, you have been tracking them during the year (see #10), and all you’ll need to do is organize them.  Besides being fun, who doesn’t like to talk about wins? It's rewarding and documenting them is important for your career.  This data can be used during reviews or if you are searching for a new opportunity.  

Be realistic and make sure you quantify wins.  If the impact can be financially quantified, that's important as dollars are a unit of value understood universally.  However, also look at significant events, and ascertain if the impact is broader than it appears.  For example, you may have replaced an end-of-life UPS with a new one.  In addition to successfully completing a major project with no downtime, you have increased availability (put a number to that, hours and potentially financial losses avoided), possibly right-sized (saving money on capital and operating expenses), and improved efficiency (additional operating savings).

2. Plan to Win. Ask yourself these questions: Have I really invested enough time to develop a comprehensive plan for 2020?  Have I finished it completely, or is it only 75% done?  Have I enlisted others (direct reports, coaches, management, financial analyst) to help, review, and/or provide advice?  Does my plan encompass all aspects of me and my team’s responsibilities?  Does it document assumptions that I can share with my management to help verify that it is aligned with my organization’s business direction?  Does my plan indicate time frames and specific dates?  Finally, have I allowed time for reviewing my plan on a monthly basis?

3. Own Your Metrics. We all recognize the importance of metrics.  However, we are often not sufficiently controlling the conversation. There are two high-level categories of metrics:

• Organization/company assigned metrics.  They are meaningful because they are dictated and should be valuable to your organization.

• The metrics you have developed and use.  Most often, your company, unless it is in the business of delivering IT, will not have metrics that provide sufficient insight into your data center operations.  You should use them and communicate their value to the organization. 

What makes a good metric?  Of course, it must be meaningful.   It also must be easy to produce and track; otherwise, it will fall by the wayside after some time.

Consider granularity of metrics to provide more insight.  You likely have an overall availability metric, but you can provide further insight by having category components such as: 

• Infrastructure – IT
• Infrastructure – Facility
• Applications
• Network

Some of your metrics should involve financial parameters.  Consider determining operational costs of computing using the metric MRC /kW (Monthly Recurring Cost). That’s a common way that colocation providers quote their services. 

Ultimately, metrics should stimulate improvement.  For example, have metrics for each major system comparing current age to normal end-of-life.  This prompts regular conversations with management about budgeting for replacements within an appropriate time frame.

4. Don’t Go It Alone. Leverage internal and external resources for your organization.  Internally, if you don’t have one already, seek at least one coach.  Coaches are often business centric; but in the data center world, you should consider a coach or two on the technical side.  We cannot know everything, so play it smart.  Additionally, make sure you leverage your team.  New or younger members of your team often provide a fresh perspective that should be considered.

From an external perspective, you likely have trusted vendors who serve your organization.  Your existing vendors are constantly receiving information from the companies they represent.  Additionally, they are out in the data center ecosystem and seeing a lot.  Maybe one of their customers is undertaking an initiative that you should consider.

AFCOM and other associations provide global and national perspectives through conferences, regional labs, and communications.  Many cities have local chapters enabling you to compare practices among your peers. Get involved, and be sure to attend Data Center World - 2020 in March. 

Tad, along with fellow DCI Board Member Bill Doty, will  present a session, "Leveraging Vendor Management to Improve Resiliency - The Hard and Soft Sides," at DCW on Monday, March 18. 

5. Stay Current. It is vital to stay current with technology and best practices.  Your management is counting on you.  This requires rigor, so think of this like your exercise program.  You must do it on a consistent and frequent basis.  Pick a few periodicals that are delivered to your inbox every day.  Set up a different email address if it helps.  For me, I find it best to spend 20–30 minutes at the start of a day.  It could be over lunch or in the evening for you.  Set an alarm and stick to it.

Consider what periodicals, blogs, etc. work best for you.  The AFCOM Journal, and Data Center Knowledge are both great valuable overall sources.  Also, consider what your CIO and CEO might be reading including WSJ, WSJ CIO, and CIO.  Make sure you are subscribed to your own company/organization’s mailing list, blog, and LinkedIn.  Set up a Google alert.

Because there are many sources of information, assign different ones to your team members and ask them to share any valuable information they learn.

6. Invest in Yourself and Your Team. Fight for training time and financial support.  Explain to management the importance of investing in this arena. Leverage HR for help in assembling your recommendations. Consider not only technical but management training.  I took a course in time management last year and discovered some areas of inefficiency in my workdays.

Invest in team-building exercises.  During the course of the year, the team will be challenged.  A more cohesive and supportive team will better endure the storms. Make sure some of these are fun!  Top Golf, or axe throwing (check with HR) are fun and popular. Be more creative than the post-work drink.

7. Leverage Data. The infrastructure supporting your data center is constantly producing data.  Some of the data you will use for your metrics, some data will define new aspects or even trends that are occurring in your ecosystem.  Inquire with your vendors as to what data can and should be captured and what data should be populating dashboards the team needs to be watching.

8. Get My Monitoring House in Order. Are your monitoring systems providing sufficient information to allow your team to fully comprehend your data center?  Is this the year that you invest in DCIM? Many vendors offer their packages in modules which lowers the cost of the initial investment.  And now you can purchase monitoring via a cloud subscription significantly reducing a possible hefty initial expense. If you already operate one, are you getting the best use out of it?  Is further training required?  Does your dashboard really reflect what it should?  

Monitoring of equipment works great if your equipment contains a NICs (Network Interface Card) or something similar.  If your equipment is not “smart,” then some level of instrumentation must be deployed.  That might involve installing CTs (Current Transformers), sensors, etc.  For example, you may want to know more precisely about the fuel level in your generators.

As camera systems continue to decline in price and become easier to install, consider deploying additional systems in key locations.  Having visibility of support systems such as generators, heat rejection equipment, switchgear, UPS and EPOs provide an additional level of security and are instrumental in incident reconstruction.

Here’s a small but important tactical task:  Make sure all the clocks on all equipment are set to the proper date and time.  It’s great that everything has a clock for monitoring events but is extremely challenging when they all reflect a different time if you have an incident.  You will be very grateful you took 30 minutes at the beginning of the year to do this.

9. Review Your Maintenance Contracts. Typically, you renew your maintenance contracts annually.  Invest some time to ensure they adequately reflect your data center’s current state.  Did physical aspects, of the data center change—quantity or type of equipment, size of data center, etc.?  Does the age of your equipment justify a change to the frequency or scope of maintenance work?  Have all your maintenance vendors submitted an updated MOP that reflects current state, physically or operationally?  Do they differentiate between minor and major PMs?  They should.

Is there an annual increase in contract cost?  Is this warranted?  Some contracts include a 3% - 5% annual increase that tracks higher than current inflation rates.

10. Document your 2020 Wins! Now, we’ve come full circle. Just as you did for 2019, document all your wins throughout the year!

The best possible scenario is for us to control our own destinies. Setting resolutions, a.k.a. goals, gives us the best shot at doing so. Best wishes for a safe, sane, and sustainable 2020!!

Tad Davies, a member of AFCOM's DCI Board, has focused on data center strategies and consults regularly with data center and executive teams for more than 30 years. He leads Bick Consulting Services, a Bick Group business unit.  He advises clients on business-centric strategy issues such as consolidation, colocation selection, build vs. buy, and cost modeling. Tad also provides advice on facility-centric strategy issues, including risk assessment, new data center programming, owner’s representation, and energy improvement.