VISTA System Management Daily or Weekly Tasks

The VISTA system, like any other database, needs to be monitored appropriately. I will mention here database related monitoring and then VISTA related monitoring. Many of the tasks that are inexpensive to check are better checked with a continuous such as Nagios or Munin.

Database monitoring

• Global growth rate
• Lock table consumption
• Daily back-up is happening daily; weekly check that the backup that was made can actually bring the system back-up
• Journal file growth rate
• Back-up and Journal file rotation (delete old ones based on business requirements)
• Database blocks integrity checks

Other system

• Check disk space for all partitions
• Check printer queues
• # of processes
• Processes consuming 100% of a single CPU that are running for more than 1 hour
• Network latency for all major network nodes in which the database is spread out; and from where it is usually accessed by users.

VISTA Specific

• Taskman performance, especially:
  • Resource Devices
  • Printer Devices
• Global growth by specific Fileman file
• HL7 messages continuous flow
• Error trap
• Taskman error trap
• Check ^TMP, ^XTMP and ^%ZTSK making sure that they are appropriately purged daily
• Check programmer mode log-ins
• Check excessive log-ins from a single location
• Check that nightly tasks ran (can be ignored as long as there are no history of problems)
• Mailman
  • Check disk usage
  • Check user's usages
  • Check queues to make sure that they are not backed up
• Check Alerts
  • How many?
  • Who isn't processing them?
  • System (errors) or Clinical Alerts?
• PIMS Reports
  • Potential duplicate patients recently entered
  • Patient movements that are incomplete (I don't know how to do that)