Proposal
Ask - Can we have more visibility into pg_stat_statements table with metrics about temp, shared and local blocks?
Use case. Why is this important?
To identify the cache hit percentage per query, this is necessary as we are in a stage of finding out the DB bottlenecks are unoptimised queries. Currently we are achieving this with having a custom binary forked from this one.
Proposal -
Flag based enablement with the binary like --collector.stat_statements.include_wal_metrics or --collector.stat_statements.include_shared_blk_metrics default to false to overcome unintentional high cardinality issues.
Proposal
Ask - Can we have more visibility into pg_stat_statements table with metrics about temp, shared and local blocks?
Use case. Why is this important?
To identify the cache hit percentage per query, this is necessary as we are in a stage of finding out the DB bottlenecks are unoptimised queries. Currently we are achieving this with having a custom binary forked from this one.
Proposal -
Flag based enablement with the binary like --collector.stat_statements.include_wal_metrics or --collector.stat_statements.include_shared_blk_metrics default to false to overcome unintentional high cardinality issues.