Bill Detwiler: If you make a mistake when upgrading a user's desktop or laptop, many times you re only affecting that single user or perhaps a handful of people that depend on that user. But if you make a mistake when upgrading a server, you could stop dozens, hundreds, even thousand of users dead in their tracks.
I'm Bill Detwiler, and during this TR Dojo episode, I'll go over five mistakes you should always avoid when upgrading a server.
Many techs have learned the hard way to avoid the first mistake on our list, which is performing an upgrade without a verified data backup. You should be making regular backups anyway, but the practice is critical when performing and upgrade.
When you power down a server, there's always a chance, slim though it may be, that it won't come back online. So even something that seems innocuous, like shutting down servers to install Windows patches, could become a disaster if the servers fail to restart, and you don't have a backup on hand.
The second mistake on our list goes hand-in-hand with the first. Not only is it critical to have a verified data back, but I also recommend you avoid upgrading servers without a way to roll back the changes.
Unplanned downtime can cost your business time and money. If an upgrade fails or breaks some critical functionality, disk images can help you quickly recover not only data but also a server's complex configuration.
Tools from companies like Acronis or StorageCraft Technology provide a universal restore option that enables recovering a failed server even to a different bare metal chassis.
If you think performing multiple upgrades at the same time is a good way to minimize server restarts, I urge you to reconsider. As this is mistake number three on our list.
I know it's tempting to add disks, replace memory, install additional cards, and upgrade software without repeatedly restarting a server. It can save time and it's generally a good idea to avoid frequent restarts.
But, if -- or when -- things go wrong a day or two later, the process of isolating the change responsible for the error is significantly more difficult when multiple changes were made at the same time. If only a single change is introduced, it's much easier to pinpoint the culprit and fix the problem.
Now, fourth on our list of mistakes, is failing to monitor the logs after restarting the server.
Remember, just because the server booted the OS without displaying errors, there could still be something happening behind the scenes that might present a problem later on.
After an upgrade, monitor the log files, error reports, backup operations, and other critical events more closely than ever.
Finally, server hardware is famously inconsistent. Manufacturers frequently change model numbers and product configurations.
So, the last mistake I encourage you to avoid is starting an upgrading without knowing for sure that the part will fit in the server's chassis.
Whether you're installing additional disk controllers, disks, memory, or other components, you should review the manufacturer's technical specifications before ordering the part.
And if possible, open the case on the actual server, or an identical one. This way, you'll be 100 percent positive that the new component will fit.
Now, these are just a few of them any issues IT pros should consider when upgrading servers. For more tips, I courage you to check out Erik Eckel's article, "10 things to remember when upgrading servers," on which this episode is based. I'll link to it from the TR Dojo blog. And be sure to let us know which issues have given you the most trouble by leaving a comment in the discussion thread.
As always, for more teachings on your path to becoming an IT Ninja, visit trdojo.techrepublic.com, or you can follow me on Twitter at twitter.com/billdetwiler.
Thanks for visiting the TR Dojo.
