Alex Williams

I’ve been working as a system administrator for the last 8 years. I do it because there’s always something new to learn. I’m constantly presented with new challenges and deep inside I like being the master of all computers systems. I think all system administrators share these traits.

On the other hand, one of the biggest problems and complaints is that we’re always on-call. When a server goes down at 3am, we have to respond. The reason is simple: we’re the only ones who know how to operate the damn thing.

The problem

I think it’s a bad idea for most companies to hire only 1 full-time sys admin. There is a specific skillset required to be a true sys admin. If only 1 person has that knowledge and can fix problems when they occur, then the company is at the mercy of that person. If that person decides to leave, they are taking all the knowledge and skills with them.

The solution

To solve this problem requires a different plan of attack. My idea is to hire 2 people with identical (or very similar) skills. Then you only give them part-time work (20 hours per week) and split the on-call duties. Here’s an example:

- Admin 1: works Mon & Tue from 9am to 5pm, and Wed from 9am to 1pm
- Admin 2: works Wed from 1pm to 5pm, and Thu & Fri from 9am to 5pm
- Admin 1: on-call Thu, Fri, Sun
- Admin 2: on-call Mon, Tue, Sat
- Admin 1 & Admin 2: rotate weekly to be on-call Wed

Each sys admin would be given the responsibility to properly document all work that is done during the daytime and keep a log of all tasks performed while on-call. When you have someone auditing your work and documentation, the chances of screwing up something are a lot lower, therefore the systems are less likely to go down during the night.

This is very advantageous for sys admins because it will also allow them to take a vacation when needed. It will give them someone with whom to share their knowledge and learn things they may not have known.

Obviously not everyone is interested by “part-time” work. Most system administrators enjoy working full-time 40 hour work-weeks. But a schedule like this, with a fair compensation and bonus for being on-call can greatly improve the life of many system administrators while helping companies gain redundancy and failover in the most important aspect of their business.

How’s that for scalability?

What do you think? Share your thoughts in the comments below.