DataGrip Connection Pool: Unbounded Max Size and Idle Handling

Question

DataGrip Connection Pool Configuration
DataGrip relies on the underlying JDBC driver to manage pooled connections. When the Max Pool Size field in the Data Source settings is left blank, the driver’s default is applied – often an unlimited value. Because DataGrip does not automatically close idle connections, the pool can grow until the database server’s connection limit is reached, causing intermittent failures for concurrent sessions.
Key uncertainties:

The absence of a hard limit means DataGrip may open an unbounded number of connections under heavy load.
Idle connections remain open indefinitely, with no built‑in cleanup or minimum idle threshold.
Connection Timeout only limits the time to establish a new connection, not the lifetime of pooled connections.

To prevent exhaustion, we need to understand:

Should DataGrip enforce a default maximum pool size when the field is empty?
Can we configure DataGrip to automatically close idle connections after a configurable period?
What mechanisms exist to monitor real‑time pool usage within DataGrip?

Tasadduq Burney · Answer

Preventing Connection Exhaustion
To prevent DataGrip from exhausting database server connections, you must explicitly define a Max Pool Size in the Data Source settings. When this field is left blank, DataGrip defers to the JDBC driver's default, which varies by vendor and may be effectively unbounded or set to a high default that exceeds your server's max_connections limit.

Recommended Configuration Steps

Navigate to Database Explorer → Right-click your data source → Properties.
Go to the Advanced tab.
Locate the Max Pool Size field and enter a conservative integer (e.g., 5 or 10) based on your server's capacity and the number of concurrent users.
To handle idle connections, check the Advanced tab for driver-specific properties such as idleTimeout, maxLifetime, or keepAlive. These are managed by the JDBC driver, not the IDE's core settings.

Technical Explanation
DataGrip functions as a client that utilizes JDBC drivers to communicate with the database. The connection pool is managed by the driver or the internal IntelliJ platform manager. Because the IDE prioritizes performance and responsiveness, it maintains idle connections to avoid the overhead of repeated handshakes. However, since DataGrip does not provide a universal "idle cleanup" toggle, the pool can grow linearly with the number of concurrent sessions or complex query windows until the server rejects new requests.

Verification and Monitoring
Since DataGrip does not provide a real-time visual gauge of the pool's internal state, you must monitor usage from the database server side. Use the following scoped commands to verify active sessions:

Database
      Verification Command

PostgreSQL
      SELECT count(*) FROM pg_stat_activity;

MySQL / MariaDB
      SHOW PROCESSLIST;

Oracle
      SELECT count(*) FROM v$session;

Note: If you are using a specific JDBC driver version or a connection proxy (like PgBouncer), please specify the driver version, as this may change the available advanced properties for idle handling.

DataGrip Connection Pool: Unbounded Max Size and Idle Handling

DataGrip Connection Pool Configuration

1 answer

Preventing Connection Exhaustion

Recommended Configuration Steps

Technical Explanation

Verification and Monitoring

Your perspective belongs here.

0 question comments

Database	Verification Command
PostgreSQL	`SELECT count(*) FROM pg_stat_activity;`
MySQL / MariaDB	`SHOW PROCESSLIST;`
Oracle	`SELECT count(*) FROM v$session;`