There are insufficient functional tests for runtime field support in Stack Rules #100738
Labels
estimate:needs-research
Estimated as too large and requires research to break down into workable issues
Feature:Alerting/RuleTypes
Issues related to specific Alerting Rules Types
Feature:Alerting
response-ops-ec-backlog
ResponseOps E&C backlog
Team:ResponseOps
Label for the ResponseOps team (formerly the Cases and Alerting teams)
technical debt
Improvement of the software architecture and operational architecture
test-coverage
issues & PRs for improving code test coverage
Edit: @ymao1 has confirmed runtime fields seem to work fine in Stack Rules out of the box, but we don't have any functional tests around this.
We should add some proper end-to-end tests verifying that this support doesn't unexpectedly break.
A Kibana dev just asked if runtime fields can be used in alerts. We'll have to check each alert to find out if they support runtime fields the way you would expect them to. For the alerting team, we should at least check out the index threshold, elasticsearch query, and geo containment alerts.
For example, for the index threshold alert, the field being compared to the threshold, and the grouping field, should both be able to use a runtime field (numeric and keyword typed, respectively).
I'm a little worried that while the executor functions don't really care whether a field is a runtime field or not, some of the alert validation DOES care (checks the types). If there's some failure, I'd expect it to be there. But we'd also want to make sure the aggregations we're doing with these fields also works as expected.
We should probably add some functional tests for this ...
The text was updated successfully, but these errors were encountered: