BangDB Server- Network Cluster (Master Slaves Model)

Benchmark - BangDB Server

Goal of the document is to measure the benchmark for server without bringing in slave synchronization or replication into the picture. This will alow us to test the performance of the server only.

The Bangdb configuration;

Following machine (commodity hardware) used for the test;

  • Model : 4 CPU cores, Intel(R) Core(TM) i5-2400 CPU @ 3.10GHz, 64bit
  • CPU cache : 6MB
  • OS : Linux, 3.2.0-51-generic, Ununtu, x86_64
  • RAM : 8GB
  • Disk : 500GB, 7200 RPM, 16MB cache
  • File System: ext4

BangDB Configuration;

  • Key size : 12 bytes - unique
  • Val size : 20 bytes - randomly picked
  • Page size : 8KB
  • Background log flush : every 50 ms
  • Buffer pool background work : ON, every 60 ms

Below are the graphs of few tests done to showcase th performance of BangDB in different scenarios






Conclusion

BangDB works very well to scale to operate on billions of records on a $400 commodity machine. The performance achieved through BangDB is also very high. The important point to note is that the performance of BangDB remains consistent and degrades gracefully under heavy load which is desireable for any high performance web scale application. Hence BangDB suits the requirement for in-memory cache or persistent data store.

Please see the comparison of BangDB with redis at Gridquorum