Content tagged with "elasticache"
Integrate the Remote Query Cache Plugin into a Spring Boot app with application.yml config and ApplicationRunner warmup. Verified with both node-based TLS and Serverless ElastiCache.
Confirmed 98.9% hit rate over 1 hour of continuous load. No timeout recurrence after 10-minute idle. Combined with AWS docs' connection reuse recommendation, Serverless + warmup is production-viable for sustained traffic.
Compared node-based+TLS to confirm the initial timeout is Serverless-specific. A warmup connection with a 5-second wait recovers CacheMonitor to healthy state, making Serverless viable for production.
Verified the Remote Query Cache Plugin in AWS Advanced JDBC Wrapper 3.3.0 with ElastiCache for Valkey Serverless. The first connection always times out, pushing CacheMonitor into SUSPECT state — a Serverless-specific issue not seen with node-based clusters.
Verified the Remote Query Cache Plugin in AWS Advanced JDBC Wrapper 3.3.0 with Aurora PostgreSQL (1M rows) and node-based ElastiCache for Valkey. A 749ms window function query dropped to 2ms. TTL-based consistency control worked as expected.
Implement same-AZ routing using Powertools getMetadata() in TypeScript. Verified caching behavior and clearMetadataCache() with real measurement data.
Hands-on verification of Lambda's new AZ metadata endpoint. Measured same-AZ vs cross-AZ latency to an ElastiCache Valkey cluster and quantified the routing benefit.
Warmup was unnecessary with plain JDBC for node-based clusters, but HikariCP's connection pool initialization makes it beneficial in Spring Boot. First request improved from 2443ms to 264ms.
Official docs don't disclose the idle timeout for Serverless. Empirically confirmed that connections survive at least 10 minutes of idle with no timeout recurrence.
The official troubleshooting guide explicitly states that the first request has higher latency. Found in the 'Reuse connections' section.
aws elasticache create-cache-cluster --engine valkey fails with InvalidParameterValue. Valkey engine requires the CreateReplicationGroup API even for single-node clusters.