Rewrite the ApproachingEnforcedSamplesLimit alert expression in the OCP documentation like "(scrape_samples_post_metric_relabeling / (scrape_sample_limit > 0)) > 0.9" (which reads as "alert when the number of ingested samples reaches 90% of the configured limit).

doc link: https://docs.openshift.com/container-platform/4.15/monitoring/configuring-the-monitoring-stack.html#creating-scrape-sample-alerts_configuring-the-monitoring-stack