Creating a filter for count of warning occurrences, but critical events are still handled on first occurrence


Hello Sensu Community!

I hope my topic was clear, but to re-iterate:
I’m trying to create a filter that will allow me to say “don’t handle warning events until there are 5 or more occurrences, but handle critical event statuses immediately.”

I tried a filter like this:

  "filters": {
    "ls_es_cluster_warnings": {
      "negate": false,
      "attributes": {
        "check": {
          "status": 1,
          "name": "eval: value == 'logstash_es_cluster_status_check_aggregate_check'"
        "occurrences": "eval: value == 5 || (:::check.interval::: * value) % 3600 == 0"

The name of the check is logstash_es_cluster_status_check_aggregate_check
I’m trying to catch "status": 1 (warning) events from being handled immediately, and to wait for 5 of them before triggering the handler, but also re-alert each hour: (:::check.interval::: * value) % 3600 == 0

I have no idea how to debug this. It’s still currently handling all events, so I must be doing something wrong :slightly_smiling_face: