Tips and Best Practices for being On-Call

Re:Tips and Best Practices for being On-Call by tt on Wednesday March 30, 2011 @04:04PM (#2388945)

It doesn't matter very much how awesome your device is. You need at least a second person who would also get the alert as a backup (assuming this is a critical system that you really care about, otherwise you wouldn't want to be on-call for it in the first place).

As the primary responder, your responsibility is to acknowledge any alert within a minute or so. Otherwise, the backup person would have to jump in.

To ensure "accountability," you may have a manager who would be a third backup and can escalate up from there.

[ Parent | on HN ]

Re:Tips and Best Practices for being On-Call by semicolon on Wednesday March 30, 2011 @05:13PM (#2389280)

Yes, a good escalation policy would have a primary responder, a backup or secondary, and then one or managers, going up the hierarchy. PagerDuty supports that and my next post will be on that topic.

Having a good device is important too though; if you sleep through or miss an alert, it may take another 10-20 minutes or so (depending on the escalation policy) before the alert escalates to the next person. This slack time could be pretty important depending on the severity of the problem.

[ Parent | on HN ]