So you want to be a geek

Generally, being a geek is not a bad job, even with the snotty people or those who are intentionally obtuse (for instance, those who ignore, three times, the very direct statement “You do not have to make any changes.”). For the most part, people are happy that you’re there, and glad that you know how to repair whatever has gone horribly awry in their technical lives. They are pleasant, civil, and just regular human beings.

Some days, though, can be trying. If you’re thinking of a life in geekdom, let me give you one of those days.

Yesterday ended for me around 2 AM this morning, because my todo list is always rather ambitious and far exceeds the number of hours available in any given day. I signed off, went to bed, and tried to sleep, only to be awakened at around 4:30 AM by our remote server monitor, telling me that there was a vast and wideranging problem and that the servers were not accessible. Crawled out of bed, contacted the NOC, to find that my suspicions were correct: one of the master breakers had tripped, and taken down one of the buss lines, including the circuit where our router was located – for those of you who don’t entirely understand that, it means that although not all of our equipment had lost power, it was in fact entirely inaccessible since our gateway was affected by the abrupt loss of power.

The standby electrician was notified, and in the meantime I had the NOC guys reset the breaker to get things back up. This was done pretty quickly, and things came right back up with very little fuss. With the exception of three servers that suddenly decided their drives needed to be checked. And the exception of one server whose primary drive killed itself due to a power surge.

While we were awaiting the arrival of the electrician to replace the breaker with a spare, I headed to the NOC to address the three recalcitrant machines and the dying server. When I got there, the electrician was there, and unfortunately, they were not able to do the swap without taking out our circuit, so down it went again, for a minute or two for the swap. When things came back up, I took care of the three servers that needed disk checks, started on the one whose primary drive had failed due to the uncontrolled reboot, and discovered that another server’s primary drive didn’t like the bounces and decided to keel over, too. So, I took care of that one as well. Four and a half hours later, I left the NOC, to return to the HQ to finish the restores on the two servers that had their drives replaced – including the one where the primary drive was a mammoth 500G SATA2 drive that will certainly be going back to the manufacturer for a warranty replace as the server is only about two months old and the drive should be within the year they give.

Back at the HQ, I checked in with the staff and began the restores. By 5:30 PM, all the restores had been completed and all the little miscellaneous (non service impacting) issues had been addressed. While those restores were going on, I started on the next batch of kernel updates on about 120 machines, answered helpdesk tickets, had some coffee, and took a break just now to type up this post. My day will probably end once again around 2 AM tomorrow morning.

Most days are not like this, thankfully. But almost every geek will have one (or more) just like it. Keep that in mind when you’re pining away for a career change, and pick something simpler. Lion taming. Stunt work. Eating fire. Something like that.