Efficiently accessing arbitrary subsets of a master sorted set in Redis

Question

Given a large sorted set with rapidly changing scores/weights, what's the most efficient way to maintain subsets and access those subsets in the master set's order?

Answer 1

You have a couple of options for this:

• Update scores in the superset and when you need to read the subset call ZINTERSTORE with subset weighting set to 0 so the result is the subset with scores from the superset.
• When updating the scores, write to the subsets as well as the superset.

The most appropriate option will depend on what you need to optimise for - read, write, memory or simplicity. How the subsets are determined and the size of the sets relative to the number of updates may also have an impact.