Nobody likes server problems. Downtime, security incidents and server interruptions not only upset clients but consume time you likely do not have. There are three simple server management steps you can take to keep your servers running great.
System Updates
A key component of good server management is system updates. Updating your server is not a difficult process. I could probably teach you the basics of how to update a Red Hat or CentOS server in a few minutes. The challenges are:
- Keeping up with updates as they arrive without notice.
- Knowing which updates are applicable to your system.
- Understanding which updates require restarts or reboots to enable.
- Fixing issues when an update breaks something.
These items are the real challenge of effective server management. Knowing when and what to patch is often more difficult than actually applying the patch. The benefits of keeping your system updated can not be overstated.
By keeping your system updated, you reduce security risks and eliminate bugs. When issues do arise, you can bypass the question, “Are you running the latest version?”, and move onto solving your service issues.
Delete It!
Too often, I see servers packed full of files, outdated sites and unused email accounts. Nearly weekly, our staff finds a server which low disk space due to an email account that has not been checked in years. I’ve seen many systems have issues due to sites that were test sites that were hacked, email accounts that were not checked, and similar problems.
If something is not critical to your current operations, delete it.
Your production system should not be used as some time vault. If you lose a client or no longer need an email account, backup the contents and then remove it from your production environment.
Don’t Ignore Repeated Issues
If you do not have a proactive server management service, then I highly recommend you keep a server diary. When you notice problems, jot them down. Over time you may begin to see patterns.
Repeated issues are often early warning signs of potentially critical issues.
I have recovered many servers where the owners simply rebooted anytime there was an issue. You should not have to reboot a Linux server nightly. As part of our server management program, we only reboot servers when there is no other alternative or a kernel update is applied. Repeated failures may indicate hardware issues, performance issues or hacking events.
If you see repeated problems on a server, investigate. If you do not have the skills to dig into the problem, find someone who does. The odd incident is not a major issue, but if at 3AM every other day your server is failing, that a clear sign of more significant problems to come.
Summary
To keep your servers running smoothly, simply:
- Keep the system updated.
- Delete what is not needed.
- Don’t ignore repeated problems.
In my 10+ years of managing servers, I’ve found these three things can keep servers running great. The form the basis of any server management program I develop.