Persistent breach detection in Kibana Universal Alerting

A persistent breach condition in a Kibana Universal Alerting rule detects a metric that stays above a threshold across several consecutive time buckets, for example CPU above 90% in all 10 of the last 10 five-minute windows. This filters out transient spikes and fires only when a problem has been sustained.

ES|QL can express this with bucket counting:

FROM metrics-*
| WHERE @timestamp >= NOW() - 50 minutes
| EVAL bucket = BUCKET(@timestamp, 5 minutes)
| STATS
    total_buckets     = COUNT_DISTINCT(bucket),
    exceeding_buckets = COUNT_DISTINCT(
      CASE(system.cpu.total.pct > 0.90, bucket, null)
    )
  BY host.name
| WHERE total_buckets >= 10
    AND exceeding_buckets == total_buckets
| KEEP host.name, total_buckets, exceeding_buckets

1. Lookback must cover all 10 buckets (10 × 5 min = 50 min)
2. Assign each event to its 5-minute time bucket
3. How many distinct buckets exist in the window
4. Count only buckets where CPU exceeded threshold;
5. null values are excluded by COUNT_DISTINCT
6. Require a full window of data before firing;
7. every bucket must have breached

The rule's lookback window must cover all the buckets you want to check. In this example, 10 five-minute buckets requires at least 50 minutes of lookback.

If any bucket is missing because the host stopped reporting briefly mid-window, total_buckets drops below 10 and the condition doesn't fire. This is a deliberate safety check: a host that went silent for one bucket is treated as "we don't have enough data to confirm persistence" rather than "breach."

If you want to allow some gaps, replace exceeding_buckets == total_buckets with a ratio or a minimum count:

| WHERE total_buckets >= 8
    AND exceeding_buckets >= total_buckets * 0.9

1. Tolerate up to 2 missing buckets
2. 90% of present buckets must have breached

Design the query so that gaps in reporting produce the behavior you want before deploying it.

• ES|QL query patterns: Browse query patterns ordered by complexity, from a basic event filter to SLO burn rate and persistent breach detection.
• Alert delay: A setting-based alternative for requiring consecutive breaches before an alert opens.