I know that when your Google App Engine (GAE) app has 0 instances running (because it has been idle for a bit), and a user requests a page, the user has to wait for the instance to boot up and do all of the instantiation which can cause the user to wait a significant amount of time.
My question is about the situation when your GAE app already has 1 instance running, but begins to experience heavy load and starts booting up a second instance.
In this case, which will happen:
-
Will a user end up having to wait for the second instance to instantiate before getting their request responded to?
-
Will no requests be sent to the second instance until it has fully instantiated, thus not making a user wait an extended amount of time?
EDIT: Its unfortunate that currently the answer to this is #1. However, there is a feature request to change the behavior to be #2. Please star this feature request( http://code.google.com/p/googleappengine/issues/detail?id=2690) to help get it to the attention of the App Engine developers
This blog post may give you some hints. Shortly speaking, the answer seems to be 1)