Metric alerts - Support equations/formulas in Metric Alerts to calculate derived metrics like failure rate
- Dominant language
- Python
- Stars
- 44.8k
- Forks
- 4.9k
- Avg merge
- 21h 10m
- Merged PRs (30d)
- 635
Description
### Problem Statement
Currently, Metric Alerts only support single metric queries. This makes it impossible to calculate percentage-based metrics like:
- Failure rate: failed_requests / total_requests
- Error rate: errors / (errors + successes)
- Success rate: successful / total
Current Workaround Gaps
Discover: Has equations but can't query spans or custom metrics
Explore: Can query spans but no equations
Metric Alerts: Works with spans/custom metrics but only 1 metric (no equations)
Use Case
Alert on API failure rate based on HTTP status codes from mobile apps (Android/iOS). Need to avoid false positives from varying traffic volumes (day/night, week-over-week).
### Solution Brainstorm
_No response_
### Product Area
Alerts
Contributor guide
Research direction
Start with the Metric Alerts area and compare its current single-metric behavior with the equations support described for Discover and the span-query support described for Explore. Define the scope for querying multiple metrics and calculating derived rates, then verify that API failure-rate alerts work across changing traffic volumes.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- observability-sre
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Quiet
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100