Skip to main content

Sequential Scans

A sequential scan (Seq Scan) happens when Postgres reads every single row in a table to find the records you asked for.

When it's OK vs. NOT OK​

  • OK: Scanning a table with 100 rows. It's often faster than an index lookup for very small tables.
  • NOT OK: Scanning a table with 1,000,000 rows. This causes massive disk IO and CPU load.

How to decide​

pgpulse tracks the ratio of index scans to sequential scans. If you see a large table with a high number of sequential scans, it's almost always a sign that an index is missing or isn't being used.

Common Fixes​

  1. Add an Index: The most common solution. Create an index on the columns used in your WHERE clause.
  2. Add Filters: Ensure you aren't selecting "all rows" when you only need a subset.
  3. Update Statistics: Sometimes Postgres chooses a seq scan because it thinks the table is smaller than it actually is. Running ANALYZE can fix this.
  4. Use Partitions: For massive datasets, partitioning can help Postgres skip entire chunks of data.