I believe it is min idle "automatic" I don't get "0"  as an option.

 

I have noticed that if Max is set to 2 and min is set to "automatic" that
sending a single request to an app results quite often in that instance
using 18-23 minutes of time on python 2.5

 

The testing is too slow, and I don't care enough to test the complete set of
possibilities.   I don't experience this on 2.7.

 

Also If I have an app that 90% of the time needs 8 instances, and Peaks at
needing 12.   I'm can't set Min Idle to 8 so that I always have 8 and 4
"idle" so that I only really get to 12 for brief times.   Setting to Min 8
Max 8 means that I peak at 20 instead of 12.   I can kind of get there with
"automatic and 4" but this seems very counter intuitive.

 

 

 

 

From: [email protected]
[mailto:[email protected]] On Behalf Of Nick Johnson
Sent: Saturday, November 26, 2011 3:20 PM
To: [email protected]
Subject: Re: [google-appengine] Re: Idle instances do not turn off
automatically. Please help

 

On Sun, Nov 27, 2011 at 3:16 AM, WallyDD <[email protected]> wrote:

Hi Nick,

Thanks for responding and looking into this.
See my response below;

On Nov 24, 8:57 pm, Nick Johnson <[email protected]> wrote:

> On Thu, Nov 24, 2011 at 5:48 PM, WallyDD <[email protected]> wrote:
> > I am having something of similar problem with instances not turning
> > off.
>
> This isn't a problem - you're not being charged for those instances.
You'll
> only be charged if demand requires sending traffic to them, in which case
> you've been saved the overhead of starting up a new instance.

I am very much being charged for these instances.


> > The resident instances do nothing and stay idle while other instances
> > serve. Not entirely sure if it is related to your issue.
>
> If you've specified a 'min idle instances' greater than 0, then this is
> behaving as documented. The point of requesting idle instances is to
handle
> sudden increases in traffic volume while more instances are being spun up
> in the background; naturally this means that they have to remain idle
while
> waiting for a traffic spike that will require them.

The second instance fires up when traffic overloads the resident
instance(s). The new dynamic instance(s) then stay on, permanently. So
I get charged for both instances, one of which does nothing.

 

You set your "min idle instances" to 1; therefore, 1 instance will be idle,
in order to handle traffic spikes. If you don't want this behaviour, set min
idle instances to 0.

 

-Nick Johnson

 


If it is behaving as documented, which part of the documentation
should I be looking at?

- sb

> -Nick Johnson

>

--
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]
<mailto:google-appengine%[email protected]> .
For more options, visit this group at
http://groups.google.com/group/google-appengine?hl=en.





 

-- 
Nick Johnson, Developer Programs Engineer, App Engine



-- 
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.

-- 
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