pg_stat_user_table doesn't exist in pg-exporter #483
Description
Hi Team,
We are using postgres 13 version, grafana - grafana/grafana:6.7.2 & the Prometheus - prom/prometheus:v2.17.0, pg-exporter - wrouesnel/postgres_exporter:v0.8.0. It's running in a docker platform.
Recently I tried to deploy the https://promcat.io/apps/postgresql dashboard
- Database Golden Signals
- Instance Health
seems like the 2 panels ( Heap Only Tuple Updates Rate & Index vs Sequential Scans ) are not showing data in panels, when i checked with Prometheus, i can able to verify the Prometheus connected with the pg-exporter and the pg-exported metrics file have pg_stat_user_tables in nested query in pg_stat_statements_blk_read_time like below
pg_stat_statements_blk_read_time{datname="postgres",query="SELECT current_database() datname, schemaname, relname, seq_scan, seq_tup_read, idx_scan, idx_tup_fetch, n_tup_ins, n_tup_upd, n_tup_del, n_tup_hot_upd, n_live_tup, n_dead_tup, n_mod_since_analyze, COALESCE(last_vacuum, $1), COALESCE(last_vacuum, $2) as last_vacuum, COALESCE(last_autovacuum, $3) as last_autovacuum, COALESCE(last_analyze, $4) as last_analyze, COALESCE(last_autoanalyze, $5) as last_autoanalyze,
vacuum_count, autovacuum_count, analyze_count, autoanalyze_count FROM pg_stat_user_tables",queryid="-1171651946716037767",server="pgs:5432",user="postgres_exporter"} 0
Not sure why the grafana panels are not showing date for those 2 panels.
Any thoughts..? OR any mis-configured here..?