Hi!

I have an important, low-traffic, internal app. There are typically 0-2 
concurrent users, the "peak" would be maybe 5-6. Plus, there are some cron 
jobs and warmup requests that are also needs to be served, but I believe 1 
or 2 resident instances should be able to handle all traffic easily.

As an experiment, what I want to completely eliminate is loading requests, 
after those 2 instances have been started. This is Java app, and the 10-15s 
additional response time caused by loading requests totally kills the 
experience. So, I'm happy to pay for the idle time, but I'd like to be sure 
that new instances are never started, and I pretty much bypass the 
heuristics and auto-scaling of the Scheduler.

Is this possible at all?

These are my current settings:

  min_idle_instances: 2
  max_idle_instances: 2
  min_pending_latency: 10ms
  max_pending_latency: 15.0s

With these, when I look at the instances I can see 2 resident plus 2 
dynamic instances, although I'm the only user currently using the app. 
(There are short cron jobs and occasional warmup requests.)

Cheers,

Aron

-- 
You received this message because you are subscribed to the Google Groups 
"Google App Engine" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/google-appengine.
For more options, visit https://groups.google.com/d/optout.

Reply via email to