How to manage nodes and users (2)

• RT: Request Tracker: excellent ticketing system, web interface and e-mail interface, very configurable

• Monitoring and notification: Ganglia saves a lot of statistics (e.g. http://status.nsc.liu.se). Nagios notifies by e-mail (or by other means) about upcoming problems. NSC gets warnings about e.g.:

  • Service/master nodes down
  • Server processes (Maui/NTPD/SMTP server) down
  • Air temperature coming out of machine too high
  • Cooling water temperature too high
  • Soon out of space in critical file systems