But then the concurrent users interfere with each others'
computations, don't they? If a user is lucky enough to hit the server
with a request when nobody else does, he'll get to utilize more memory
(and potentially arrive at a different answer) than if he hits the
server when 10 other people do, and have all of their computations
compete for memory. That does sound like an unfortunate trait in a
On 2008.04.14., at 15:42, Randall R Schulz wrote:
> As far as it being single-threaded, when running as a CLI command,
> it is
> (JVM threads notwithstanding), and in that setting it's no big deal if
> OOM handling is disorderly. The user can simply re-run the program
> different limits (memory or operation count). But when it runs in a
> servlet container, there are potentially other threads executing
> competing user request as well as whatever other threads the servlet
> container uses to do its things.