Proactive sysadmin tasks

Question

What things can be done around the network in the unusual time when a sysadmin is not busy.

Answer 1

In no particular order:

• Freak out because you have spare time
• Test your backups
• Check your Anti-Virus
• Check your updates (WSUS)
• Update network documentation (eg: Touch up your SysAdmin Wiki)

and possibly the most important...

• Answer lots of questions on ServerFault!!!

Answer 2

Do you have centralized logging yet?

Get to it! Set this up ASAP.

I had all my windows and UNIX machines sending event log / syslog entries to a central syslog server, and a perl script in cgi-bin parsing this, highlighting scary things, filtering out the mundane, and displaying other things for categorization.

The ability to read through the logs of ~100 machines (or more, once your filters and categorization get better) in a minute or so is huge for staying in touch with what's going on in your LAN.

Answer 3

work on automation solutions to improve reporting of network/ad/etc to sysadmin. (eg report dcdiag status from all dcs to you every morning). warning : results might make you have even more "free" time

Answer 4

Documentation, documentation, documentation. It's not fun but its a necessary evil that often goes overlooked. Good documentation is worth its weight in gold especially when you're out of the office on the beach with your family and someone tries to call you to do a relatively simple task. With proper documentation you can have a jr. level employee or operations personnel tackle those tasks using that documentation. Plus it gives management the warm and fuzzies that if something were to happen to you (or whomever is responsible for that system) that support can be transferred with relative ease.

Answer 5

I like to interact with the end user during slow periods. Often this will reveal an issue you were not aware of. Then they are happier because they are getting something that bothers them fixed. Most importantly you are creating a good relationship. You are also letting them know you are a real human and not some kind of troll living under the server room floor.

Answer 6

• Clean up Active Directory. Invariably, you wind up with empty Groups, empty OUs, and test AD accounts.

• Backup your GPOs

• Print out your GPOs

• Start/Finish that desk manual.

Answer 7

• Be up to date with the latest news/blogs
• Check if your backups work
• Check if any machines are lying idle and get them back to good use
• Take inventory of all software's and hardware's this will give a good base for upgrade plans
• Check if your software's have critical patches
• Learn a good scripting language : My Choice shell or Perl or python or Ruby and on windows PowerShell

Answer 8

Do some security checking and hardening of your infrastructure

• Open ports discovery (using nmap)
• Simple (or not) vulnerability analysis (using nessus)
• Checking server configurations from a security point of view
• Firewall rules rewview
• etc.

After knowing the security state of your infrastructure you can start making changes to improve it in two ways:

1. Removing the vulnerabilities (the so-called hamster wheel of pain if you don't do number 2.)
2. Identifying the root causes and implement measures to avoid them in the future

Answer 9

There's no such thing as "Free time", there's just time to do all of those things that dropped to the bottom of the todo list. I think if I had free time, I would wonder where all the users went and why I didn't get the memo about the site closing for good.

Answer 10

Identify servers that are candidates for decommissioning. Begin investigative work necessary to prove or disprove that the server should be decomissioned.

In a similar vein, identify systems that are in need of upgrade, and start making the case for that.

Answer 11

If I had spare time, I'd probably try to work out what my cause of death was.