Summary: 20 instances, 20 unique

Text	Count
// TODO: Keeping query validation simple for now, only term aggregations have full support for the "group by" on the	1
// TODO: Can refactor this method to use Sets instead which can cleanup some of the categorization logic (like getting completed alerts)	1
// TODO: Remove "AmazonES_all_access" role?	1
// TODO: Can this be updated to just use 'Instant.now()'?	1
// TODO: Can change the parameters to use ctx: BucketLevelTriggerExecutionContext instead of monitor/trigger and	1
val bucket: Map<String, Any>? // TODO: Should reduce contents to only top-level to not include sub-aggs here	1
// TODO: This is only used to retrieve the retryCause, could instead fetch it from the bulkResponse iteration below	1
// TODO: Check how this behaves during rolling upgrade/multi-version cluster	1
// TODO: Update BucketLevelTriggerRunResult.alertError() to retrieve error based on the first failed Action	1
// TODO: What if the failure cause was not TOO_MANY_REQUESTS, should these be saved and logged?	1
* TODO: Deprecate API for a set of new APIs that will support this APIs use cases	1
// TODO: This is largely a duplicate of runAction above for BucketLevelTriggerExecutionContext for now.	1
// TODO: This method has grown very large with all the business logic that has been added.	1
// TODO test this part by passing sub-aggregation path	1
// TODO: Since a composite aggregation is being used for the input query, the total bucket count cannot be determined.	1
// TODO: Should triggerResult's aggregationResultBucket be a list? If not, getCategorizedAlertsForBucketLevelMonitor can	1
// TODO: If/when multiple input queries are supported for Bucket-Level Monitor execution, aggTriggerAfterKeys will	1
// TODO: Figure out a way to use SearchTemplateRequest without bringing in the entire TransportClient	1
// TODO: can we use one instance of the script for all buckets? it should be stateless?	1
// TODO: This should be limited based on a circuit breaker that limits Alerts	1