Engula AutoBench (Test Report)#

1. Design Goals and Comparison Methodology#

• Goal: Compare the Get/Set throughput and latency of Engula 2.1 and Redis 7.2 across different value sizes and CPU core counts.
• Metrics: QPS (throughput), P50 (average latency).
• Comparison methodology:
  • Single-machine comparison with persistence disabled, focusing on in-memory path and protocol processing performance.
  • I/O thread count dimension: io-threads ∈ {1, 2, 4}.
  • Data scale, request distribution, and concurrency model are fixed, so results are reproducible.
  • Value sizes cover multiple tiers: 8B, 32B, 128B, 512B, 1KB, 2KB.

2. Test Environment#

2.1 Hardware and System#

• Instances: Alibaba Cloud ecs.c9i.xlarge / ecs.c7a.4xlarge / ecs.g8y.2xlarge
• CPU: 8 vCPU ~ 16 vCPU
• Memory: 32 GB
• Operating System: CentOS 7.9

3. Engula AutoBench#

This test uses the Engula AutoBench tool. For a detailed introduction to the tool, see the documentation: Engula AutoBench Tool Usage.

4. Test Results#

ecs.c9i.2xlarge#

• CPU: Intel® Xeon® Granite Rapids * 8
• Memory: 16 GB
• Operating System: CentOS 7.9

ecs.c7a.4xlarge#

• CPU: AMD EPYC™ MILAN * 16
• Memory: 32 GB
• Operating System: CentOS 7.9

ecs.g8y.2xlarge#

• CPU: Yitian 710 * 8
• Memory: 32 GB
• Operating System: CentOS 7.9