I’ve spent most of the past four years talking to sysadmins.  Sysadmins come in many different varieties.  Some sysadmins work by themselves, and they either know enough about everything under their purview or they know how to work a search engine well enough to fake it.  Other sysadmins are specialists, and they know one thing backwards and forwards, and they know a lot about what hooks up to their one thing, and they usually know at least something about many other things.

When I’m asked to describe sysadmins, I often describe them as people who are very good at solving problems.  I still think that this is true.  Now I think it’s the side effect of something more central.

Sysadmins are tinkerers.  They might be buttoned-up tinkerers who always have a backup of whatever it is that they’re working on so that they can revert to a known good state.  They might be inveterate tinkerers who will have an idea and just try it out, and rely on their tinkering skills to be able to get themselves out of a jam.  In either case, they’re tinkering to get it to work better.  “Better” can mean anything from using less power to integrating with another application to giving more meaningful alarms.

Sysadmins tinker not just for the satisfaction of tinkering.  They tinker to make their lives easier.  A friend told me that the best sysadmins are lazy.  They don’t want to get woken up in the middle of the night, they don’t want to have to repeat the same thing.  They tinker so that they have less to worry about and less work to do.

The form that this tinkering takes can be of many types.  It could be learning all of the internals of an application so that they can better integrate it into their infrastructure.  It could be scripting so that they can automate something that is important to them, or annoys them, or that they just wish that they didn’t have to do.  It could be playing with the latest and greatest technology to see if it really is as great as the hype to see if it meets their needs.

The sysadmin gets an idea and tries out solutions.  They tinker with it to see if they can get it right.  They poke and prod at it, and try it from different angles, and see what works.  They don’t decide on a solution (although they might have a solution decided for them) and go out and read up everything on it and get trained on that solution.  They try out the solution to see if it really does fit their needs.  If their tinkering results in something really useful, they might invest in something like training, or they might just continue tinkering and figuring it out on their own.