There are a lot of variables here that make it very hard to generalise, for example:
- how big is the data (and keys)?
- is the time being dominated by serialization/deserialization? (if so, I can recommend a few other serializers :p)
- what does the test code look like? is it actually optimal?
- is the server load representative? (a single-threaded test loop over an otherwise silent database is good for repeatable numbers, but doesn't actually tell you much about real load)
Personally, assuming my servers are local to each-other, and the payloads aren't big: I'd expect lower numbers than that - based on my real world usage.
Do you have a test rig I can look at?
Marc
--
You received this message because you are subscribed to the Google Groups "Redis DB" group.
To unsubscribe from this group and stop receiving emails from it, send an email to redis-db+u...@googlegroups.com.
To post to this group, send email to redi...@googlegroups.com.
Visit this group at http://groups.google.com/group/redis-db.
For more options, visit https://groups.google.com/d/optout.
Oh, and yes: that scenario is one of the things that is usually pretty expensive for SQL server and cheap for redis.
Finally, you haven't indicated your redis config - is it configured for persistence, for example? (Usually not required in a cache server)