On this page is a list of checks you should do as a System Administrator for your systems. I like suggestions if I've forgotten anything, or tips how to handle these checks. The point is really to write a best practice procedure. This is not a complete list, it's just the checks I do for my company and checks I did for previous companies, completed (?) with the checks my (ex-)colleagues do.

The first part is really the checks. What do you check and how frequently. The second part is how to make this into a procedure. The last part is really important if you're company is looking for certification or want to prepare for certification.

What do you check and how frequently?

Considerations:

• You should always check the backup, and when a backup is not necessary the backup line should state : “Backup: not necessary.”
• You should always check the UPS communication
• Antivirus (AV) should always be checked on Windows machines, but can be neglected for *nix

<servername>

• Syslog (/var/log/messages)
  • See SYSLOG
• Uptime
• Diskspace % Free
  • /
  • Data volume
• Backup
• UPS

<servername>

• Syslog (!not enabled by default!)
  • See SYSLOG
• Uptime
• Diskspace % Free
  • /
  • /data
• Backup
• UPS

<servername>

• Server health
  • Remote Manager
• Uptime
  • monitor.nlm
• Diskspace
  • SYS
  • DATA1
• Backup
• UPS

<servername>

• Event Viewer
• Services
• Uptime
• Diskspace
  • C:\
  • D:\
• Backup
• UPS

• edir synchronisatie
  • ./dsrmenu.sh
• ndsstat

• edir synchronisatie
  • iMonitor

• GW errors to GW administrator
• GW Check Reports
  • POA
• GW log files
  • POA
  • MTA
  • WebAccess
  • GWIA
• backup

• Check the /var/log/oracle/<version>/<oraclesid>/bdump/alert_<oraclesid>.log
  • Errors in this logfile start with “ORA-”

I like to work locally since it's faster, so I shouldn't forget to copy my data to the network:

• Create backup local data
• Copy local data to network

Websites need special attention as well:

• Create backup from website
• Check comment on website (if possible)
• Check email from website

I'm a documentation freak so I say, once a week:

• Check to see if all work of the week is documented
• Check documentation from colleagues

• As a system administrator you have to deal with prejudices so
  • Clean up your desk once a week
  • Read and reply to your email
  • Update the status of your open incidents/issues/tickets/calls

• Backup data paths in nodes
  • Note the date when you last done this

• Make sure all clients and the management server are up to date
  • Note the date when you last done this

• Check backup overview and procedure
  • See Template Backup Data Paths
  • See Backup Procedure
  • Let all responsible key employees also recheck
  • Note the date when you last done this

• eDirecory health check
  • See eDirectory Health Check

• Check year backup
• Check if there's been a full restore this year, if not perform one