PagerDuty Help
We use mulitple tools to monitor our systems, and send us alerts that we should inspect something. When you see such an alert, you are expected to make a best effort response.
Your first step is to read the alert. What is the message telling you? If the alert is not helpful, please take note of that as an action item to improve in the very near future. Vague or confusing alerts waste everyone’s time. In an ideal world, we only ever get alerted for things that demand our immediate attention, and which we can actually remedy. It’s a long road to get to that ideal world! Take every chance you get to increase the accurancy and usefulness of alerts.
In the event that you can’t resolve the problem on your own, you’ll need to get help. The way to get help is to use PagerDuty. PagerDuty has a robust scheduling system with overrides and escalation policies. By using PagerDuty – and only PagerDuty – you ensure you reach the right person to help you!
| In Slack | On PagerDuty |
|---|---|
Type /pd trigger to open the Slack workflow: | Go to https://banno.pagerduty.com/ and click “New Incident” at the top right. |
PagerDuty works on services, so you may need to triangulate between an application and a service. In Digital, we use the Banno Domain Ownership document to help us find who owns what.
Every team should have a generic service named “Team
Please do not assign this page to a specific person. That defeats the purpose of schedules and schedule overrides!
Give a brief but informative title and description. “Help” and “Service Down” and similar generics are not helpful. Remember: you’re sending a message to someone’s phone. They may be eating a meal, or fast asleep. You want to give them enough useful information – description of the problem, what Slack channel to join, and anything else immediately relevant – that they can arrive reasonably prepared to help you!
Do Not
- direct message someone for help
- call or text someone for help
- @channel or @here for help
Also, do not forget to use good judgement. This process is the standard, but there may be legitimate reasons to explicitly violate this standard. Do the right thing.