SharQ - An Open Source Rate Limited Queueing System - PyCon India 2014

SHARQ is a flexible, open source, rate limited queuing system. Based on the Leaky Bucket Algorithm, SHARQ lets you create queues dynamically and update their rate limits in real time.

Demo content from the talk:
https://github.com/sandeepraju/PyCon-India-2014

More Details:
http://sharq.io
https://github.com/plivo/sharq-server
https://github.com/plivo/sharq

Transcript

1. sharq An Open Source Rate Limited Queuing System
2. Sandeep Raju Product Engineer @ Plivo
3. THE PROBLEMS Unpredictable traffic spikes on cloud infrastructure Gracefully handling 3rd party resource limits
4. Request  Graph  of  the   HTTP  Server   Request  graph  of  the   Service  Cluster  
6. Solution: Introduce a Queue
7. Solution: Introduce a Queue
8. Enter SHARQ •  Ability to create queues dynamically to ensure fair queueing •  Ability to change the rate limits in real time •  Have constant flow (not burst)
9. HANDS ON DEMO
10. The internals Based on the leaky bucket algorithm rigid output pattern at the average rate, no matter how bursty the input is
11. The internals Python, Flask, Gevent, MsgPack Lua for atomic operations Redis as the broker
12. SHARQ IN PRODUCTION Highly Available
13. SHARQ IN PRODUCTION Horizontally Scalable
14. Roadmap of SHARQ Enqueue Dequeue Finish Interval Metrics Job Status Max Retries Richer Analytics SHARQ Client Library Your Suggestions...
15. Check out sharq.io for more information & documentation
16. Q&A Check out sharq.io for more information & documentation