This is a quick note
1. Query Conditions Not Matching the Index
- The index is not on the columns being queried.
- The query does not use the leading columns of a composite index.
- The query is not written in a way that takes advantage of the index's order.
- The query uses functions or operations that prevent the use of the index (e.g., using functions like LOWER() on the column in the query).
- Data type mismatches between the column and the query can prevent index usage. Ensure that the data types match exactly.
2. Query Planner's Cost Estimates
- PostgreSQL's query planner estimates the cost of using the index and might decide that a sequential scan is cheaper.
- The statistics for the table are outdated or inaccurate, leading the planner to make suboptimal decisions. Running ANALYZE can help update these statistics.
3. Index Selectivity
The index is not selective enough. For example, if a column has many duplicate values, scanning the index might not save much time compared to a sequential scan.
Top comments (0)