ability to utilize more than 1 cpu.

It's unlikely to happen in the short- to medium-term.

What about the 10 concurrent request limit? Do you think a Go instances
could safely handle more then 10 in flight request? If so how may? I would
like to create a feature request.

I created a test app that simply held requests open for 10 sec. On App
Engine a new instance was added on the 11th request, while a single
instance on Heroku handled 200+ request before I received the first error.
While Heroku instance are most likely beefer then AE I still think that AE
can do better.

It seems that this limit is imposed to protect python and java apps more
then Go, is that correct?

Sorry to hijack the thread, but it seems related.

That's not clear. The true limit of an instance is obviously highly
dependent on the workload of the app; some could handle thousands, and
some struggle with 10. 10 seems to be near a reasonable sweet spot
that is safe for almost all apps, but also reasonably effective for
concurrent apps.

Feel free to file a feature request on the issue tracker to make that
configurable. I doubt we will change away from 10 as a default.

I could believe that. Try doing some work in the app handler and the
balance might change.

The general mechanism is to protect all of them, but it is conceivable
that we could some day make it tunable.

Dave.

Thank you Dave for your quick responses. I can see how 10 makes for
a responsible default.

For anyone else who is interested in allowing their instance to handle more
than 10 request, I have created a feature requests - please star it:
http://code.google.com/p/googleappengine/issues/detail?id=7927

Thanks,

Kyle

thanks kyle, starred