Ah, thanks for the explanation. I'm actually exploring Redis for storing timelines!
Twitter's Redis-backed timeline storage is here, if anyone is interested: https://github.com/twitter/haplocheirus
Twitter's Redis-backed timeline storage is here, if anyone is interested: https://github.com/twitter/haplocheirus
This operations are useful every time you want to use Redis for caching timelines, as the idea is: if this user is already in cache (at least a single item exists) then push against the cache, otherwise not.