My current apps use tiny amounts of memory, so that would be good for me 
too. :D However, to be fair, apps that consume huge amounts of CPU should 
have some form of quota too.

The problems with the idea of paying for instances include, but are perhaps 
not limited to:

1. No incentive for Google to make their scheduler better (the more 
instances the scheduler creates, the more money Google makes and the more 
extra the customers have to pay).

2. Unless carefully defined, the idea of an 'instance' is an arbitrary 
concept in terms of how many instances are crammed into each server, how 
much effective CPU time and other resources each instance gets in reality, 
and how much overhead is included. A term that Google can change their 
definition of at their own whims. That can be opaque as hell. And here 
again, if the customers pay for all the overhead caused by the GAE 
infrastructure software, then there is no incentive for Google to improve 
the performance of that code.

3. Unclear and difficult for customers to estimate how many instances they 
will have to pay for each month even when they know the average load of 
their application, since the spawning of new instances is determined by 
Google's code, not the application code.

-- 
You received this message because you are subscribed to the Google Groups 
"Google App Engine" 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?hl=en.

Reply via email to