Parent ticket: #101016
Summary
Background: #108115 (comment)
During index bootstrapping there can occur certain situations related to network conditions:
-
Timeouts, for example when network or ES cluster are under load. Currently we have 20 minutes timeout for installing common ES resources shared between all indices + 20 minutes timeout for installing index-specific resources for each index separately (e.g. for .alerts-security.alerts). Total 40 minutes.
- During these 20-40 minutes the rules will be blocked on attempting to write alerts and will be hanging. It will look like "going to run" status in the Rule Management table in Security and no logs or other messages.
-
Errors, like network errors or errors from Elasticsearch.
- In this case, errors will be re-thrown as exceptions; the rule status will change to "failed" and there will be some Kibana logs available.
Do we need to build a better UX around that?
Parent ticket: #101016
Summary
Background: #108115 (comment)
During index bootstrapping there can occur certain situations related to network conditions:
Timeouts, for example when network or ES cluster are under load. Currently we have 20 minutes timeout for installing common ES resources shared between all indices + 20 minutes timeout for installing index-specific resources for each index separately (e.g. for
.alerts-security.alerts). Total 40 minutes.Errors, like network errors or errors from Elasticsearch.
Do we need to build a better UX around that?