https://github.com/sonots/redis_vs_q4m

benchmark to evaluate redis vs q4m as a queue engine
https://github.com/sonots/redis_vs_q4m

Last synced: about 2 months ago
JSON representation

benchmark to evaluate redis vs q4m as a queue engine

• Host: GitHub
• URL: https://github.com/sonots/redis_vs_q4m
• Owner: sonots
• Created: 2014-10-27T18:54:06.000Z (about 10 years ago)
• Default Branch: master
• Last Pushed: 2015-03-22T08:12:53.000Z (almost 10 years ago)
• Last Synced: 2024-10-13T08:08:43.173Z (3 months ago)
• Language: Ruby
• Size: 227 KB
• Stars: 5
• Watchers: 0
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
redis_vs_q4m

============

benchmark to evaluate redis vs q4m as a queue engine

# RESULT

Benchmarking enqueue/dequeue 10,000 messages. 

|                         | enqueue (sec) | dequeue (sec) |

|-------------------------|---------------|---------------|

| q4m (--with-sync=yes)   | 1.886356      | 3.476402      |

| redis                   | 0.460380      | 0.271107      |

| redis(to_json)          | 0.573746      | 0.402243      |

| redis(persistence)      | 0.517071      | 0.310961      |

In throughputs (#/sec) form:

|                       | enqueue (#/sec) | dequeue (#/sec) |

|-----------------------|-----------------|-----------------|

| q4m (--with-sync=yes) | 5301            | 2876            |

| redis                 | 21721           | 36885           |

| redis(to_json)        | 17429           | 24860           |

| redis(persistence)    | 19339           | 32158           |

CONCLUSION: Redis won. But, please note that redis' rpop does not have rollback functionality. 

# Machine Spec

```

CPU Xeon E5-2670 2.60GHz x 2 (32 Cores)

Memory  24G

Disk    300G(10000rpm) x 2 [SAS-HDD]

OS CentOS release 6.2 (Final)

```

Q4M

```

q4m-0.9.13-1.mysql_5.6.20

```

Redis

```

redis-2.8.17

```

# conf

Q4M

* [my.cnf](q4m/my.cnf)

Redis

* [redis.conf](redis/redis.conf)

* [redis.aof.conf](redis/redis.aof.conf) for presistence

# ToDo

* benchmark in multithreads

  * redis is single thread, mysql (q4m) is multi thread. So, q4m would overcome redis

* create docker container or AMI to make it easy to rerun the benchmark