on-call

Incidents, stack traces and postmortems.

2 public promptsSearch within #on-call

Prompts tagged #on-call

Read this stack trace from service name (language / framework). In three sentences: what failed, the most likely root cause, and the first thing I should check. Then list up to three other possible causes, ranked, each with the log line or metric that would confirm it. If the trace is cut off or ambiguous, say what extra information you need instead of guessing. paste stack trace
Arjun Mehta2,240 uses
Turn this incident timeline into a blameless postmortem. 1. Summary: what users saw, for how long, how many were affected 2. Timeline: in time zone, one line per event 3. Root cause: technical, never "human error" 4. What went well, what didn't 5. Action items: owner, due date, and whether each one prevents or detects If the timeline has gaps longer than 15 minutes, point them out. Timeline: paste timeline
Daniel Okafor690 uses

Tags with fewer than 3 public prompts aren't shown to search engines yet.