Radovan Jablonovsky <radovan(dot)jablonovsky(at)replicon(dot)com> wrote:
> In usual load there are not much pressure on memory, but it is
> possible to have all clients start using heavy reports. They are
> valid requests and could consume all memory.
Your clients will get their results back faster if you can arrange
some way to queue these sorts of requests when they get beyond some
reasonable limit. You might be able to do that using a connection
pool, or you might want to create some sort of job queue which
releases a limited number of such jobs at a time; but I guarantee
that every single person to submit a job to such a queue, including
the last person, will get their results sooner with such queuing
than turning loose a "thundering herd" of requests which puts the
system into swapping. I guarantee it.
-Kevin