Parent ticket: #101016
Summary
The indexing logic of Rule Registry (RuleDataService and RuleDataClient) is not covered with tests, not well tested and its quality and stability is unknown.
We need to cover it with integration and unit tests, checking:
- initial index bootstrapping
- subsequent index bootstrapping with the same parameters
- subsequent index bootstrapping with changed parameters (new mappings added/removed, settings changed, custom ILM policy set, etc)
- simulated race conditions between "kibana instances" bootstrapping the same index
- simulated race conditions between "plugins/solutions" bootstrapping different indices
- simulated race conditions between "kibana instances" bootstrapping multiple indices
- error handling of responses from Elasticsearch
- error handling of network errors
- error handling in data processing (our own code)
- writing a lot of documents to a single index
- writing a lot of documents to multiple indices
- load on ES due to many calls to
getWriter or getReader
- etc, think of anything else
Parent ticket: #101016
Summary
The indexing logic of Rule Registry (RuleDataService and RuleDataClient) is not covered with tests, not well tested and its quality and stability is unknown.
We need to cover it with integration and unit tests, checking:
getWriterorgetReader