Check and double-check

What's the best way to make sure your network doesn't get blindsided by a disaster? According to Ed Bott, checking and double-checking key systems is critical. Find out what safeguards TechRepublic members suggest to keep your network running smoothly.

What do airplane pilots, NASCAR drivers, and network administrators have in common? In all three cases, a crash can have serious career-limiting consequences. The best way to thrive and survive in any of those high-pressure fields is to check and double-check key systems regularly. In my most recent Microsoft Challenge, I passed along a request I found on the Technical Q&A boards. The administrator of a small- to medium-size network (up to 500 users) wants to assemble daily, weekly, and monthly lists of general responsibilities for everyone on the IT team. Fellow TechRepublic members came through with some great suggestions.

TechRepublic member Juan DeMarco offered the best explanation of why a formal checklist is so valuable: "I am a contractor for many small- to mid-size businesses, and a lot of the problems they call me for [can be prevented by] easy precautionary measures." His list, in fact, is so clear and straightforward, he routinely hands it to clients and recommends that they incorporate it into their daily, weekly, and monthly routines. Other TechRepublic members, including DRDON and winwalker, skipped the introductions and simply posted their lists.

If you've got a complex network, I recommend that you read over the complete, unedited responses to this Challenge. I've excerpted the best suggestions in this executive summary:

  • Backup, backup, backup. "I can't emphasize that enough," said DeMarco. He also cited two customers who lost millions of dollars because they skipped the next item.
  • Check the backup! Make sure the backing program is running smoothly and that all appropriate files are being backed up. Store the tape in a secure location.
  • Perform routine maintenance on user accounts. Add new accounts and monitor shares, security groups, and disk quotas. Disable old accounts and manage account policies.
  • Monitor event logs and services. Don't forget routers and firewalls.
  • Perform routine disk maintenance on servers. Make sure you have sufficient free space on all servers. ("Don't let users run out of space," cautioned TechRepublic member DRDON.) Defrag and check disks, as long as you can do so without having to reboot.

  • Check performance on the network and on key servers. DeMarco said, "I use Performance Monitor to check CPU, Memory, Network Utilization, and HDD Access at 10-to 15-minute intervals during business hours throughout the week." A quick review of these logs can help you quickly spot when something's not right.
  • Purge temporary files on servers.
  • Perform routine disk maintenance on client PCs. Defrag and check hard disks to identify problems before they become crises.
  • Implement new policies, permissions, logon scripts, or scheduled script modifications.
  • Audit the network for unauthorized changes, inside and outside.
  • Check for essential system updates and patches. Check Microsoft's Downloads page and Windows Update Corporate Site for service packs and hotfixes.
  • Install the latest antivirus definitions.

  • Perform a full backup and send a copy for offsite storage.
  • Check for software updates. Don't forget BIOS revisions. If you've standardized on a small number of system configurations, this task should be relatively easy.
  • Check hardware performance. Think of this step as the equivalent of a tune-up and oil change. "Check the hubs, switches, and routers for collisions or other network anomalies," DeMarco recommended. "Find out quickly that your network is having problems before they bring you to your knees." On a high-performance, mission-critical network, you might want to perform this task weekly.
  • Audit security settings. Pay special attention to service account and admin passwords.
  • Gather and review statistics. This is a good opportunity to evaluate performance and procedures and identify any areas for upgrades and changes.

That should be enough to get any network administrator started. Thanks to everyone who contributed to this week's column.

Here's Ed's new Challenge
I just experienced every Windows user's nightmare—a sudden, serious, data-destroying hard disk crash. Luckily, I was able to recover all my data from backup tapes, but I wonder whether I could have done a better job of anticipating this crisis before my disk began making horrible grinding sounds. How do you keep tabs on disk performance? Do you have a set of procedures and third-party tools you use to keep your disks running at peak efficiency? I have only two requirements:
  1. Any tests must be able to run without requiring a reboot.
  2. Any utilities must be fully compatible with Windows 2000.

If you think you’ve got a winning disk disaster prevention plan, click here to tackle this week's Microsoft Challenge and take a shot at earning 2,000 TechPoints.