A couple other things. Make sure you warm up your instance before doing your test. If you try that loop as your initial hit to the server than the instance has to start up which could also cause you to timeout. Also, run the test a few times. With Java's JIT thresholds your going to get a speed bump once it has run several times.
On Thu, Jul 28, 2011 at 4:30 PM, Stephen Johnson <[email protected]>wrote: > You'll probably want to look at and test backends where you can pick your > processor speed to see what kind of performance you can get. I think the > front-end instances are configured similar to B1's, but i can't remember if > I saw that somewhere or am basing it on the fact that in the new pricing > model on-demand hours are priced at $.08, so maybe it is similar to B2, > maybe someone else knows. Also, it was mentioned a while back that a B8 > can't do any faster calculations on a single thread than a B4 so you'll > probably see similar results if you test your computations on a B4 and B8. > But once again, not completely sure about that assertion. The upside to a B8 > from a B4 is, from what I gathered, increased memory and if using Java then > can have multiple requests running at same time on different threads, it's > just that a single thread can't do better than B4. > > B1 128MB 600MHz $0.08 B2 (default) 256MB 1.2GHz $0.16 B4 512MB 2.4GHz$0.32 > B8 1024MB 4.8GHz $0.64 > > > Stephen > CortexConnect > http://cortexconnect.appspot.com > > On Wed, Jul 27, 2011 at 5:00 PM, SudoKU <[email protected]> wrote: > >> I am a new comer to the Google App Engine. I was testing whether I can use >> Google app engine for a real-time calculations of our geophysical model. The >> idea is that scientists can calculate / run my model online and plot the >> results rather than running natively on their computer. I am testing the >> free version of google app engine. To test the speed of computation, I wrote >> the following loop in the Java servlet. I wanted to know how fast Google App >> Engine can compute my codes. >> >> >> >> for (int i = 1 ; i < 100000 ; i = i+1) >> >> { >> >> for (int j = 1 ; j < 100000 ; j = j+1) >> >> { >> >> for (int k = 1 ; k < 100000 ; k = k+1) >> >> { >> >> int h = i + j + k; >> >> } >> >> >> >> } >> >> >> >> } >> >> >> While the local deployment in my laptop (I am using eclipse & google >> plugin - localhost:8888) completed the calculation in 8 seconds, google app >> reports error after a few seconds. If I remove the outer loop, Google App >> engine complete the calculation in about 4-6 seconds. While the actual >> computational burden of my model can be less than that of the above codes, I >> was a disappointed by the fact that Goggle App Engine runs slower than my >> laptop. Is this because of the free version ?. Will the speed increase, if I >> pay for App Engine use ? >> >> >> Thanks >> >> >> >> -- >> You received this message because you are subscribed to the Google Groups >> "Google App Engine for Java" group. >> To view this discussion on the web visit >> https://groups.google.com/d/msg/google-appengine-java/-/VpNTC_2FI3wJ. >> To post to this group, send email to >> [email protected]. >> To unsubscribe from this group, send email to >> [email protected]. >> For more options, visit this group at >> http://groups.google.com/group/google-appengine-java?hl=en. >> > > -- You received this message because you are subscribed to the Google Groups "Google App Engine for Java" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/google-appengine-java?hl=en.
