5 Redis Use Cases with Gur Dotan - Redis Labs

[@wnwca]

Thanks for such an excellent talk. Covering five key use cases in a well organized way.

[@KevinGenus]

Fibonacci use case! FINALLY. Thank you!!

[@lmaoiwaslikelmao1132]

Gur Dotan is the most Warcraft Orc name I've ever heard in real life

[@DawidOhia]

Great presentation! I wonder what is the averege hit/miss ratio and the size (in number of entries) of the cache in this Dataroma scenario.

[@sunwonjhung9919]

Great Explanation! I have a question, how do you invalidate the caches? Do you use TTL?

[@Redisinc]

TTL is a great candidate and our main recommendation for an easy invalidation policy.

[@rohitchatterjee2327]

Excellent slides very easy to follow

[@osamaa.h.altameemi5592]

Fantastic talk, thx a ton.

[@ערןאוצפ]

Great presentation, 2 questions 1) how do u corelate a cached query result with a new incoming query (meaning recognizing the same query again) 2)this is more of a clarification then a question, A query is taken from the queue and then attempts to accuire a lock for that specific tannet. If that tannet is already at it's capacity the query is re queued , potantioally it might be busy the next time around and the next time.. etc. And then the query is thrown away.

[@dandymcgee]

For #1: It's probably just a hash of the query text (maybe auto-formatted in some way to remove minor, inconsequential differences like whitespace?). Either way, there are likely cases where the "same" query would have a cache miss due to a minor difference in the query text. This is almost certainly irrelevant enough to ignore. Re #2: He said there's a max retry count. After that limit is hit, the query is indeed discarded. Presumably the client would then be notified of this via the pub/sub in the form of a "query failed" message rather than a query result.