Leverage connection pooling and multiplexing for performance and scalability
summary
Demonstrates that Oracle Database Resident Connection Pooling reduces host memory and scales far beyond dedicated servers: a cited example with 5,000 clients drops memory from ~42 GB (dedicated) to 2.8 GB (shared) and ~1 GB with DRCP. Shows Python Flask + Oracle driver using server_type="pooled", dbms_connection_pool APIs, and ApacheBench (100 requests) where purity=self yields 0.453s vs purity=new 0.769s. Notes DRCP supports tens of thousands of connections, is enabled by default on Autonomous Database, and exposes V$CPOOL_* and DBA_CPOOL_INFO for monitoring.
AI-written summary · about
chapters (10)
- 0:00Introduction to DRCP
- 1:02Scaling database connections
- 4:14Benefits of DRCP
- 6:16DRCP architecture overview
- 7:39Operational mechanics
- 11:29Use cases and memory impact
- 13:36Configuring DRCP
- 18:26Client-side optimization
- 22:06Monitoring DRCP
- 23:12Practical implementation
related talks
33mData handling techniques in Python for Database
39mDatabase Connection Pool Sizing - Demystified!
20mElephant by the pool: PostgreSQL connection poolers overview
29mFuture-proof your applications using RAC
24mSimplify application development with Oracle Real Application Clusters
42mEfficient Data Handling Techniques in Python for Oracle Database
13mOracle Globally Distributed Database
13mA Deep Dive into the DB Connection Pool