AnsweredAssumed Answered

Producer Consumer process design

Question asked by matroyd on May 24, 2012
Latest reply on May 25, 2012 by gera
An application has a **Producer processes**, and another **Consumer processes** (not threads) running simultaneously. The consumer process needs to access the data as soon as Producer process is done generating it. The producer moves on to generating new data once it is finished with one set. Producer process is very CPU intensive, whereas Consumer process is I/O intensive and spends a lot of time waiting.

How to go about designing this elegantly using MapR?

One of our early proof of concepts used MapR via NFS and demonstrated above usecase using lock files (O_EXCL).

Outcomes