Investigate logs and possible incident causes
Provide the incident time, symptoms, logs without secrets and recent changes. Gryphon can reconstruct the timeline and help test possible causes.
Example: errors after a timeout change
Some requests fail after a configuration update. Sanitized logs and the diff are available.
- Reconstruct the timeline and affected operations.
- Compare timeouts, retries and error handling in code.
- Propose checks for hypotheses and measures to prevent recurrence.
What you get
A timeline, ranked hypotheses and a verification plan for the incident responder.
Collect the incident data
Include the environment, affected operations and timestamps with a time zone. Add errors, metrics and recent diffs.
Test the hypotheses
Compare events with the code and configuration. For each possible cause, ask what supports it and what check could disprove it.
After recovery
Write up the incident, prepare a fix and add a regression test. Production commands require the access and permission you grant.
Try it on your own task
Open Web and describe the task. Connect Desktop for files and commands on your computer.