|
|
|
# Wiki for Redis Project
|
|
|
|
|
|
|
|
## Steps:
|
|
|
|
1. Get Redis running (http://redis.io/)
|
|
|
|
- run server locally
|
|
|
|
- play around with it using command-line client (redis-cli)
|
|
|
|
- try programming with another client (such as the python, ruby, or php client)
|
|
|
|
2. Try running Retwis (Twitter clone) off of the Redis instance
|
|
|
|
- PHP: https://github.com/antirez/retwis
|
|
|
|
- Python: https://github.com/pims/retwis-py
|
|
|
|
3. Run *sharded* Redis ("Redis Cluster")
|
|
|
|
- configure Redis to split keys among multiple shards
|
|
|
|
- can still just run these shards locally
|
|
|
|
4. Simulate load on the cluster
|
|
|
|
- collect some data about timing/latency/throughput
|
|
|
|
- see if we can actually see it slowing down
|
|
|
|
5. Figure out how sharding works
|
|
|
|
- Look at the code, find where the sharding happens
|
|
|
|
- Figure out what configuration options there are
|
|
|
|
- Start to determine where we would make our changes
|
|
|
|
6. Hotspot detection
|
|
|
|
7. Replicate hot keys (read-only)
|
|
|
|
8. Implement split keys (allow commutative operations) |
