Bench-Markingยท

Memcached, Apache Ignite, and Redis ๐Ÿš€

Built a 3 caching diffrent systems with python and enclosed in containers.

The natural world is teeming with creatures of all shapes and sizes, each with its own unique set of adaptations and behaviors. While many animals pose little threat to humans, there are some that command respect and caution due to their deadly capabilities. From apex predators to venomous insects, let's explore the 10 most dangerous creatures on Earth, culminating in the ultimate threat: humans.

Overview ๐Ÿ”

This report compares the performance of three popular caching systems: Memcached, Apache Ignite, and Redis. The comparison is based on write latency, read latency, and connection duration metrics.

Memcached

1Average Write Latency~25.15ms
2Average Read Latency~43.48ms
3Median Connection Duration (50th percentile)~0.041s

Apache Ignite

1Average Write Latency~35.62ms
2Average Read Latency~31.43ms
3Median Connection Duration (50th percentile)~0.025s

Redis

1Average Write Latency~2.99ms
2Average Read Latency~2.64ms
3Median Connection Duration (50th percentile)~0.0026s

Summary

Redis demonstrates superior performance across all measured metrics: ๐Ÿ†

  • Lowest write and read latencies (under 3ms for both operations)
  • Fastest median connection duration (~2.6ms)

Apache Ignite shows mixed results: ๐Ÿ”„

  • Better read performance than Memcached
  • Slightly worse write performance than Memcached
  • Faster connection times than Memcached, but slower than Redis

Memcached exhibits the slowest overall performance: ๐Ÿข

  • Highest latencies for both read and write operations
  • Particularly slow read operations compared to other systems
  • Slowest connection times among the three

Conclusion ๐Ÿ“Š

Based on these benchmarks, Redis emerges as the most performant caching system, followed by Apache Ignite. Memcached shows the slowest performance overall.

the redis results didn't seem realistic to me, so I decided to run some more benchmarks to verify the results. guess what? the results were different. it was even more faster. heres an official way to benchmark redis using redis-benchmark cli tool.

valley

traveler

Note ๐Ÿ“

These results are based on specific benchmark tests and may vary under different conditions or workloads. I used containerized environments for each system to ensure consistent testing conditions.

Resources

Resources and links for the project.

Source Code

Github repo for the project.

> cdย ..

Balaji ยฉ 2020-2025