SQLExporterScrapingLimit #

Meaning #

Alert is triggered when Prometheus takes too long to fetch the SQL exporter metrics

Impact #

The monitoring system is degraded. SQL exporter does not collect SQL metrics, alerts cannot be triggered.

Diagnosis #

1. Check CPU and IOPS usage of the PostgreSQL server

   An overloaded server may have difficulty collecting metrics.

2. Look at PostgreSQL Server logs to identify long running queries.

   You may need to enable log_min_duration_statement to identify which queries are long to be executed.

Mitigation #

1. Identify and kill heavy queries

   How to terminate queries?

   SELECT
       pg_terminate_backend(pid),
       usename,
       datname,
       application_name,
       client_addr,
       client_port,
       state,
       wait_event_type,
       wait_event,
       state_change,
       query
   FROM pg_stat_activity
   WHERE pid in ('<replace_with_pids>');
   

Additional resources #

n/a