All mappings need support for named regex capture groups to pull values out of the metric name OR match product standards
We see many cases where names are unique per instance, even though this is an anti-pattern for Prometheus.
Eg of the metric:
infinispan_mycache1_latency_total{
infinispan_mycache2_latency_total
infinispan_mycache3_latency_total
We should be able to create a metric mapping:
prometheus.metric.sets=infinispan
prometheus.metric.infinispan.metric_list=infinispan_(?<cache_name>[a-z0-9]*)_latency_total
prometheus.metric.infinispan.apm_metric_path=infinispan|(host)|(cache_name)|(statistic)
We should also be able to use the metric set name in the metric path, and possible use some metric values as variables for other metric sets (which implies needing order of querying).
Another example:
vendor_cache_manager_default_cache_ecomm_api_ear_ecomm_api_war_rpc_manager_number_xsite_requests{...}
vendor_cache_manager_default_cache_ace_api_ear_ace_api_war_rpc_manager_number_xsite_requests{...}
vendor_cache_manager_default_cache_integration_api_ear_intg_api_war_rpc_manager_number_xsite_requests{...}
An enhancement request was already raised in order to have it in our monitoring portfolio, it will be evaluated and put in our backlog to be deployed soon.
To follow up on the status of this request, please raise a ticket on https://support.broadcom.com/ and refer to this knowledge article in order to get the status.