Summer breeze makes IT fine, blowing through sysadmins in my mind
Taking a break from those never-ending help tickets
Comment Summer is here and IT bods can rejoice because if users aren’t already on their annual two-week holiday then those remaining – the ones with kids – will be soon once schools break up in mid July.
It’s the time of year when the heavy lifting side (physical or virtual) of infrastructure upgrades can be done without pesky users complaining about email being slow or raising help tickets.
From my experience this is the time of year when I’d often see pallet upon pallet of new IT kit getting dropped off, be it servers or desktops. Even some “cash strapped” government agencies seem to be doing very well judging by the sheer amount of kit that comes in.
Also around this time of year, projects get unbunged and the IT guys and girls can get their new kit landed, unboxed and into the server room and start to implement the hardware and software.
Summer tends to be a popular time to do such upgrades in larger companies as there are no change freezes like those you see at Christmas for instance, where there might be a party and alcohol-induced lull in work, but you can’t take advantage by undertaking upgrades lest you incur the wrath of the Change Management group.
It is not only physical equipment that comes through but also projects that have been gathering momentum after 1 April start to emerge and land on the desks of the IT department. The IT department becomes a hive of activity but how can you make the best of the time window available?
The tasks that admins can get on with really depend on the type of business involved. All said and done, it comes down to the question of are you a “shutdown friendly” type of business. Some larger factories and educational establishments tend to fall into this environment and have an exceptionally seasonal client base.
In academia, very few students tend to hang around uni on summer break and even fewer will want to use the IT infrastructure as beer, part-time jobs and the opposite sex hold a much more understandable appeal! Those organisation’s administrators can use this time to ready for the fresh onslaught of undergraduates wanting to login and start using (or abusing) the infrastructure come September.
So what should you be looking at for July and August?
Server hardware refreshes
Obviously as hardware gets older, the platform needs to be migrated to newer, faster and usually cheaper infrastructure than the original hardware. Lower system usage levels means less time is required to do the migration due to that extra bit of available capacity.
It also means that time that would otherwise have been used supporting user requests can be better used to push these hardware refresh projects out the door and make their admins lives easier.
This isn’t restricted to just servers either. It is prime time to replace storage arrays and other back-end components. The number of times that management have come to me and said: “Here is two hundred grand. It needs to be spent and invoiced before 1 April," are too numerous to count.
Every year, without fail. This lump-sum purchase has sat around until we could find the time to install it – that often came during the summer lull.
Interestingly, a lot of companies I have spoken to use this time of year to perform migrations from local infrastructure to cloud. As we know, cloud is getting bigger and a lot more people are looking to migrate there.
As to cloud being a good idea, that depends on a lot of factors. It is also not just applications that get migrated but also email, in huge volumes. Cloud email solutions are becoming ever more popular, and are often seen to relieve a lot of problems from the administrators workload associated with local, on premise email systems.
I expect to continue to see an ever-increasing amount of cloud mail migrations to online mail platforms such as Office 365 and Google Apps for Business platforms.
Having seen and dealt both good migrations and horrific ones I offer a word of warning to sysadmins: the boss may well have made the decision to migrate, but it is your ass on the line. You need to make absolutely sure those accounts that are migrated to the cloud work as expected.
This is why it is useful to have at least a smattering of users around to verify the procedure works for the rank and file. You may well have tested with “test” accounts but there is no substitution for testing with real honest-to-goodness accounts.
The last thing an admin wants is hundreds of potentially knackered accounts meaning wasted migration time, bandwidth and resources, not to mention billable hours of sysadmin time to fix the situation, assuming you are automating the migration (and if not, why not?)
Desktop refreshes are straightforward stuff, when done right. Before doing the rip and replace make sure you have a working procedure for migrating the user data and any user profile data across to the new desktops.
I'm sure I don’t need to tell you, but make sure you have good backups.
Major software upgrades
Software upgrades are the most noticeable summer disruption for users and the business in general. It is more about the end user training and hand holding that is required (to ensure it goes smoothly) that's the issue. Users hate change. Disruption means less effective workers until they get up to speed.
This is perhaps the reason that most companies avoided Windows 8 like the plague. It was too different. That is, however, a different argument. For the first few weeks there will be many questions so make sure that the helpdesk are prepared to answer the questions.
A pro tip is to give the users a two-page cheat sheet on how to use the new version of the software. Software upgrades tend to chew up an awful lot of business cycles so it is imperative that you have the infrastructure and personnel to make it work.
A well-planned summer upgrade can save you a lot of hassle and issues with users who don’t understand that IT upgrades need to happen. It provides a time when work tends to be at a low point and you are able to get stuff done.
But the important point to note that what ever you do now needs to be not just planned and executed well but extensively tested, too. Users coming back to a smoking heap of IT will not be impressed, and neither will the boss! ®