I take issue with this:

On Sep 8, 2011, at 6:07 AM, Tim wrote:

>  - An instance starts when either
>     - it is needed to serve a request (there are no IDLE or FREE-IDLE 
> instances available subject to max-pending-latency) - starts in ACTIVE state

I have seen cases where an instance starts for no apparent reason.  In my case, 
I'm bumping along with just one instance most of the time.  Every now and then, 
the scheduler starts a second instance, even though the first one is still 
responding quickly and hasn't handled a request in a few seconds.

My theories are:

- There is a race condition, data sync, or timing issue, which causes the 
schedule to sometimes think an instance is busy when it isn't; or,

- There are heuristics in the scheduler which predict an increase in load, and 
act proactively (such as a request coming in from a new IP address)

My details are in the thread "Scheduler Bug or Inscrutable Scheduler 
"Feature"".  No response from our google overlords on that one.

Another case was reported as well, although he had hundreds of instances 
running, so it would be a lot harder to grok the state of the system in his 
case.

-Joshua

-- 
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