SearchDataCenter.com Advisory Board
A comprehensive collection of articles, videos and more, hand-picked by our editors
As unlikely as it seems to some IT folks, there will come a time when users are out of the office and there are no urgent tickets at the help desk. During those times, IT project planning plays a key role. The SearchDataCenter Advisory Board shares a few project ideas for data center staff to tackle while the users are away.
Pete Sclafani, CIO, 6connect
Downtime is perfect for small projects that provide lots of value.
SSL certificate expirations and domain registrations: For some reason these tend to not be tracked until something is down. Just doing some lookups for key domains and putting calendar reminders (or just renewing early) will solve a future problem.
The other good downtime exercise is "security stuff." When things are quiet and working as expected, do some general audits. For example, verifying permissions for users/groups for various systems. These tend to not be touched unless required by a personnel change. But given the activity you likely see over the course of the year, taking a step back and looking at the security landscape can be good preventative maintenance.
Internal user survey: Outbound communication is a challenge when you are busy, but setting up a survey can be done any time, then sent out to the appropriate audience. Identify one or two aspects of the business that could use some feedback and put surveys together. That way you have a pipeline of communications that you can draw from during the rest of the year to reach out, and it also gains you a channel for communicating relevant information to your group. Here's an example survey idea related to BYOD policies and use cases: Ask what devices users own and how they want to interact with the business on the devices. Tie this to a BYOD policy initiative and it's an easy way to get feedback and also see what expectations users would have with a policy.
Robert McFarlane, principal, data center design, Shen Milsom & Wilke
Most people aren't going to like my answer, but its data center documentation that is always in need of work. Documentation is the thing most IT people like to do least, which makes it even more important to get it done when there's an opportunity for IT project planning.
If you don't have a good data center infrastructure management (DCIM) system, then keeping up with your documentation is critical. You can avoid the shutdown of a bunch of equipment from a circuit overload just by knowing what's in each cabinet and on each circuit before you add the next piece of hardware. If you have any intention of implementing DCIM -- as more and more IT teams do -- you're going to need that detailed inventory to do the implementation. Might as well get it done while it's quiet.
You can also use this time to get trash out of the server room that never belonged there in the first place, and look at the filters on servers to see if they're accumulating dust. Visible dirt will reduce air flow, and also indicates that dirt is entering your data center that should be isolated outside. If dust is visible, a good, careful cleaning would be worthwhile, or you might identify examples of contamination that justify hiring a good data center cleaning service. And you can also alert your facilities staff that the source needs to be found and closed. Foot wipe mats at the entrances may be all that's necessary, so use the time to get them requisitioned.
Also, look at places where blanking plates haven't been replaced or holes in floor tiles haven't been closed and air is leaking through. In other words, this is the day to take care of things that folks know should be done but never have enough time to do. This is the day when that excuse disappears!
But the most important part of a quiet day is to avoid things that really don't need to be done. In other words, abide by the old adage "If it ain't broke, don't fix it!" Maintenance is certainly necessary, but it can also introduce problems that weren't there before. Maintenance should be scheduled, with contingencies in place in case something does go wrong. That's true even in data centers with redundant infrastructure that have been designed for concurrent maintainability. Haphazardly maintaining critical infrastructure equipment, just because "there's time today," is a good way to create a disaster.
Sander van Vugt, independent trainer and consultant
Simple and in one word: upgrades! During production weeks, no maintenance windows are available to do proper upgrades of the server operating systems.
Now that everyone has left for holiday, it's easy to schedule a serious maintenance window. One caveat: You risk working all summer long while everyone else is enjoying the beach!
Clive Longbottom, co-founder, Quocirca
An average data center should use a quiet day for best practice housekeeping. Put tools in place that allow you to carry out a full inventory of fixed as well as less-connected (e.g., BYOD, mobile equipment) hardware and all software assets and Software as a Service (SaaS) usage.
This will allow some heavier stuff to be considered at a later date -- for example, how to rationalize and consolidate the use of applications and SaaS within the organization and how to create a secure information strategy. It may well be an eye opener to find out exactly what users are doing with their own kit.
Inventorying these devices may show that you have up to 20% more servers than you thought -- with all the associated costs around that. It will also enable a closer look at how virtualization and private and public cloud can provide the organization with an optimized platform. So keep it simple; don't change anything too much while people are away, but ensure that you are really helping the organization at a strategic level going forward.
Robert Crawford, systems programmer
We seem to lurch from one crisis to the next, barely keeping our systems and heads above water. Along the way we make notes about cool little things we could do to improve our professional lives and availability numbers. Every year, the list gets longer while maintaining the status quo gets more difficult. Maybe with a couple of weeks to rub together, we could finally create some of those tools that will, as management constantly asks, help us work smarter, not harder.
Do some exploring. Take a few system dumps and figure out what's new with CICS/TS 5.1. I would catch up on my Assembler coding and try out some of the new instructions IBM adds with each new generation of hardware. And break a few things -- within reason -- to find out what makes them tick. This type of exploratory work is never wasted, but proves useful at the times you least expect it.