You should probably do this using a Graph Database, eg: Neo4j. You can
solve the problem with one query, instead of actually performing
calculations: but this obviously would cost your application another
database, and some learning possibly. But its always cool to learn new
tools.
Neo4j has R
my advice on splitting this 2 processes is not use http, use child_process
and use the other as a unix executable. if later you end up splitting into
machines use a simple queue over http to accomplish this. this way you will
not mix the solutions.
On Mon, Mar 17, 2014 at 4:31 PM, Trevor Norris w
On Sunday, March 16, 2014 10:47:22 AM UTC-7, Andy Zelinski wrote:
>
> I am hoping for someone to point me in the right direction (examples,
> info) on how best to handle the need for some processor heavy computations
> that might best be executed with C++ math libraries.
>
> is this best handl
You can use a child_process (or a cluster of child processes) to do this.
But I'd actually advise to isolate that logic into a small service (either
written in node or maybe in something that is more tailored towards CPU
heavy apps, e.g. C++ or Java) and use for example HTTP to call it. This
al
