On Tue, Apr 6, 2010 at 2:12 PM, Steve <sjh_cassan...@shic.co.uk> wrote: ... > Should I assume that it isn't common practice to write updates > atomically in-real time, and batch process them 'off-line' to increase > the atomic granularity? It seems an obvious strategy... possibly one > for which an implementation might use "MapReduce" or something similar? > I don't want to re-invent the wheel, of course.
I think this is indeed a common and common sense approach. Traditionally this has actually been done by even simpler systems, like writing log files on hosts, beaming them up to a central location, and processing aggregates that way. But key/value stores can be used to reduce latency that is the problematic part of simple log-based aggregation. -+ Tatu +-
