ConnectionPoolExhausted cloud resource_error ai_generated true

RDSProxy: Connection pool exhausted for database 'mydb'

ID: cloud/aws-rds-proxy-connection-pool-exhausted

Also available as: JSON · Markdown · 中文
85%Fix Rate
87%Confidence
1Evidence
2024-05-12First Seen

Version Compatibility

VersionStatusIntroducedDeprecatedNotes
RDS Proxy 1.0 active — — —
Aurora MySQL 8.0 active — — —
Aurora PostgreSQL 15 active — — —
RDS for MySQL 8.0 active — — —

Root Cause

RDS Proxy's maximum connection pool size for the target database is reached, often due to too many concurrent connections or long-running queries holding connections open.

generic

中文

RDS Proxy针对目标数据库的最大连接池大小已满，通常是由于并发连接过多或长时间运行的查询占用连接未释放。

Official Documentation

https://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/rds-proxy.html#rds-proxy-connection-pool

Workarounds

  1. 85% success Increase RDS Proxy's MaxConnectionsPercent and SessionPinningFilters in the proxy target group: `aws rds modify-db-proxy-target-group --db-proxy-name my-proxy --target-group-name default --connection-pool-config MaxConnectionsPercent=90,SessionPinningFilters=["EXCLUDE_VARIABLE_SETS"]`. Also increase database max_connections: `aws rds modify-db-instance --db-instance-identifier my-db --db-parameter-group-name custom-params` with max_connections=500. 
    Increase RDS Proxy's MaxConnectionsPercent and SessionPinningFilters in the proxy target group: `aws rds modify-db-proxy-target-group --db-proxy-name my-proxy --target-group-name default --connection-pool-config MaxConnectionsPercent=90,SessionPinningFilters=["EXCLUDE_VARIABLE_SETS"]`. Also increase database max_connections: `aws rds modify-db-instance --db-instance-identifier my-db --db-parameter-group-name custom-params` with max_connections=500.
  2. 90% success Optimize application code to close connections promptly using connection pooling libraries (e.g., HikariCP with maxLifetime=300000 and idleTimeout=600000). Set connection timeout in AWS Lambda: `import pymysql; conn = pymysql.connect(..., connect_timeout=5, read_timeout=30)`. 
    Optimize application code to close connections promptly using connection pooling libraries (e.g., HikariCP with maxLifetime=300000 and idleTimeout=600000). Set connection timeout in AWS Lambda: `import pymysql; conn = pymysql.connect(..., connect_timeout=5, read_timeout=30)`.

中文步骤

  1. Increase RDS Proxy's MaxConnectionsPercent and SessionPinningFilters in the proxy target group: `aws rds modify-db-proxy-target-group --db-proxy-name my-proxy --target-group-name default --connection-pool-config MaxConnectionsPercent=90,SessionPinningFilters=["EXCLUDE_VARIABLE_SETS"]`. Also increase database max_connections: `aws rds modify-db-instance --db-instance-identifier my-db --db-parameter-group-name custom-params` with max_connections=500.
  2. Optimize application code to close connections promptly using connection pooling libraries (e.g., HikariCP with maxLifetime=300000 and idleTimeout=600000). Set connection timeout in AWS Lambda: `import pymysql; conn = pymysql.connect(..., connect_timeout=5, read_timeout=30)`.

Dead Ends

Common approaches that don't work:

  1. 80% fail

    RDS Proxy pool size is limited by the database's max_connections; increasing proxy percent alone does not help if the database limit is reached.

  2. 90% fail

    Restarting clears the pool temporarily, but the application will quickly exhaust it again if connections are not closed properly.

  3. 75% fail

    Multiple proxy endpoints share the same underlying database connection limit; this only distributes load but does not increase total pool size.