Hi, we are trying to locate the impact of:
1 ) unused metrics on prometheus query performance on long term
- unused labels on prometheus query performance on long term
but cant seem to find, answer to it. From resources on the internet it is clear that high cardinality used metrics and labels gonna have low performance on long term and sometimes recommend to avoid using them if not neccessary, this makes to have another question
- does used high cardinality metrics/labels has bad performance only for expressions which refers them or generally they lower the performance of all queries on cluster wide?