Has anyone noticed a dramatic difference in Cold Start times when using 
different AppIDs?

I recently changed the AppID for a project that I'm working on and there 
was a noticeable reduction in the *cold* start time.

Intrigued, I decided to perform an experiment:

   - I loaded the application on 4 different AppIDs all of which were *HRD*.
   - Using *tools.pingdom.com* I loaded the home page for each of the 
   AppIDs using the *http://*.appspot.com* domain and recorded *wait 
   (yellow bar) *time for the HTML page.
   - I, then, *shutdown* each of the instance and repeated the test.

In this limited sample the *fasted* AppID loaded the application in an 
average of *1.35* secs, while the *slowest* averaged *3.55* secs

The Application is using:
  runtime: *python27*
  threadsafe: *false*

I also performed the test on *warm* instances, but the results did *not*seem to 
indicate a correlation.

I also tested the fasted and slowest AppIDs with a similar application 
written in Go, where AppID did *not* effect the cold start times.

A little more information on the AppIDs:

   - The *fasted* AppID was created yesterday
   - The *slowest* was created many months ago.
   - The *slowest* AppIDs were also the ones containing a dash(-), but 
   that's probably just a coincidence.

Is this slowness due to older hardware, or is this somehow related to the 
python27 runtime?

Can anything be done for these slow AppIDs? Is it best just to delete them?


Thanks,


- Kyle


https://docs.google.com/spreadsheet/ccc?key=0Apr2o_vtfYo3dFNPMDQ2ZHU1T2ZwcmJIOHNQYU9EWXc

-- 
You received this message because you are subscribed to the Google Groups 
"Google App Engine" group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-appengine/-/CH89M2Tb9ekJ.
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?hl=en.

Reply via email to