Read performance is a factor of how many store files you have, which is a factor of how many flushes you end up with (approximately). If you put it in 1 RPC call or 2, it should not matter too much. But like JD said, fewer RPC calls = better performance.
-ryan 2010/7/22 Héctor Izquierdo Seliva <[email protected]>: > > > El mié, 21-07-2010 a las 09:39 -0700, Jean-Daniel Cryans escribió: >> So you would buffer edits going to the same row? Unless you have your >> own write-ahead-log, you'd likely lose data on node failure. >> >> But WRT your question, 5 cells with different timestamps is as costly >> to store/query as 5 cells with the same timestamp. The major >> difference is that the former case will require 5 RPCs, instead of 1. >> >> J-D > > That makes it clear for me. I can sort the information I have to process > so I can put all the data in a row in a single call, but it takes a > little bit of time to do this. I don't care about write performance as > much as read, hence my question about what was better. > > Thanks guys! > >
