How to Monitor Performance in Supabase

Supabase includes a built-in performance advisor that analyzes your database and suggests improvements. Go to Dashboard → Database → Performance Advisor (or Indexes). The advisor scans your tables, identifies missing indexes based on query patterns, and suggests which columns to index. It also flags unused indexes that waste storage and slow down writes. Review each suggestion and apply the ones relevant to your workload.

The pg_stat_statements extension tracks execution statistics for every SQL statement. It is pre-installed on all Supabase projects. Open the SQL Editor in Dashboard and query the view to find your slowest and most frequent queries. Look at mean_exec_time for average duration, calls for frequency, and total_exec_time for cumulative impact. Focus on queries with high total_exec_time first — they have the biggest overall impact.

1-- Find the 10 slowest queries by average execution time
2SELECT
3  calls,
4  round(mean_exec_time::numeric, 2) AS avg_ms,
5  round(total_exec_time::numeric, 2) AS total_ms,
6  rows,
7  substr(query, 1, 100) AS query_preview
8FROM pg_stat_statements
9ORDER BY mean_exec_time DESC
10LIMIT 10;
11
12-- Find queries with the most cumulative time (biggest overall impact)
13SELECT
14  calls,
15  round(mean_exec_time::numeric, 2) AS avg_ms,
16  round(total_exec_time::numeric, 2) AS total_ms,
17  substr(query, 1, 100) AS query_preview
18FROM pg_stat_statements
19ORDER BY total_exec_time DESC
20LIMIT 10;
21
22-- Reset statistics (useful after adding indexes to measure improvement)
23SELECT pg_stat_statements_reset();

Once you identify a slow query from pg_stat_statements, use EXPLAIN ANALYZE to see exactly how PostgreSQL executes it. This shows whether the query uses indexes (Index Scan) or reads every row (Seq Scan), how many rows each step processes, and where time is spent. Look for sequential scans on large tables — these are prime candidates for new indexes.

1-- Run EXPLAIN ANALYZE on a slow query
2EXPLAIN (ANALYZE, BUFFERS, FORMAT TEXT)
3SELECT * FROM orders
4WHERE user_id = 'some-uuid'
5ORDER BY created_at DESC
6LIMIT 20;
7
8-- If you see "Seq Scan" on a large table, add an index:
9CREATE INDEX idx_orders_user_id ON orders USING btree (user_id);
10
11-- For queries that sort, a composite index helps:
12CREATE INDEX idx_orders_user_created ON orders (user_id, created_at DESC);

Go to Dashboard → Reports → API to see request metrics for your Supabase REST API. This panel shows total requests, response times (p50, p95, p99), error rates, and request distribution by endpoint. Pay attention to the p95 and p99 latencies — these represent the experience of your slowest 5% and 1% of requests. If p99 is much higher than p50, you likely have intermittent slow queries or connection pool exhaustion.

For production applications, Dashboard monitoring is not enough — you need alerts. Supabase supports log drains that forward logs to external services like Datadog, Grafana Cloud, or any HTTP endpoint. Go to Dashboard → Settings → Log Drains to configure a destination. Once connected, you can create alerts for error rate spikes, slow query thresholds, and connection pool saturation.

1-- Check current connection usage (important for monitoring)
2SELECT
3  count(*) AS total_connections,
4  state,
5  usename
6FROM pg_stat_activity
7GROUP BY state, usename
8ORDER BY total_connections DESC;
9
10-- Check database size
11SELECT pg_size_pretty(pg_database_size(current_database())) AS db_size;
12
13-- Check table sizes for growth monitoring
14SELECT
15  schemaname,
16  relname AS table_name,
17  pg_size_pretty(pg_total_relation_size(relid)) AS total_size,
18  n_live_tup AS row_count
19FROM pg_stat_user_tables
20ORDER BY pg_total_relation_size(relid) DESC
21LIMIT 10;