Limits and rules pertaining to New Relic alerts and applied intelligence:
Matched data points per minute, per account (learn more)
Web app response percentiles per account
Targets (product entities) per condition
5000 targets for NRQL conditions 1000 targets for non-NRQL conditions
Thresholds per condition
1 Warning or 1 Critical
1 Warning and 1 Critical
Incidents per issue
Incidents beyond this limit will not be persisted.
Incident search API: page size
1 page (less than or equal to 25 incidents)
1000 pages (25K incidents)
Only use the
Initial limit 1000
Workflow filter size
4096 characters per workflow
Notification channels (Legacy):
2500 channels per type
Exception: No limits for user channels
If this limit is exceeded, you won't be able to create or update conditions for the impacted account until the rate goes below the limit. Existing alert conditions are not affected.
You can see your matched data points and any limit incidents in the limits UI.
To understand what conditions are leading to the most throughput, you can perform a query like:
FROM NrAiSignalSELECT sum(aggregatedDataPointsCount) AS 'alert matched data points'FACET conditionId
Some tips on optimizing your matched data points:
- If you're using sliding windows, note that this can significantly increase the number of data points. To lower the number of data points, you can use a longer aggregation duration.
WHEREclauses to scope down the amount of data being alerted on. Using
FACETcan produce more efficient alerts in some cases.
- Combine similar alerts. If you have several alert conditions that are similar, consider grouping them together with combined filters.
To request a limit increase, talk to your New Relic account representative.
Note that using sliding windows can significantly increase the number of data points. Consider using a longer duration of Sliding window aggregation to reduce the number of data points produced.