In effect it helps cover the cost of the scheduler infrastructure
itself - that handles launching instances. Also things like before an
instance can serve your application, it must download all the your
code from a 'repositiory'. There may be millions of physical machines
hosting instances, but each of them does not keep the code for every
application 'in case' on their local disk. (for example)
Which is probably why a respawned instance doesnt incur the cost,
because the code is already on the local disk. After the instance has
been shutdown, and not used for 15 minuts, the code could be flushed
from disk. meaning next time that application is served on that
machine, the code will be fetched again.
(I dont know this, just what seems logical based on what observed from
the 'outside')
On Fri, Sep 2, 2011 at 3:56 PM, GR <goo...@gmail.com> wrote:
> If an instance is available almost the instant you need one, how is it fair that you charge for a 15 minute startup time?
>
> --
> 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/-/TS8c1KBLkJgJ.
> To post to this group, send email to google-a...@googlegroups.com.
> To unsubscribe from this group, send email to google-appengi...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/google-appengine?hl=en.
>
>