Hi all,
So, I have an important, but low traffic internal app. Typically 0-2
concurrent users, the "peak" would be max 5 users. Considering the warmup
requests and some cron jobs, 1 or max. 2 instances should be sufficient to
easily serve these.
What I'm trying to achieve is that using Application Settings, I want to
have 2 resident instances (F4) and no loading requests after the 2
instances were started. This is a Java app, and loading requests add 10-15s
to the response time, totally killing the experience. I want to bypass the
heuristics and auto-scaling in the Scheduler and always run on those 2
instances, not more, not less. I am ready to pay for the idle time, but I
don't want users to face those 15s responses.
My current settings are:
<min-idle-instances>2</min-idle-instances>
<max-idle-instances>2</max-idle-instances>
<min-pending-latency>10ms</min-pending-latency>
<max-pending-latency>15.0s</max-pending-latency>
Even with these, I can see the 2 resident instance plus 0-2 more dynamic
ones. It really drives me crazy. Currently it is 4 instances and it is only
me using the app (plus the occasional warmups and cronjobs)...
Any suggestions? (I know that the long discussions about the Scheduler in
this group, but I haven't seen this exact question asked ever.)
--
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 google-appengine+unsubscr...@googlegroups.com.
To post to this group, send email to google-appengine@googlegroups.com.
Visit this group at http://groups.google.com/group/google-appengine.
For more options, visit https://groups.google.com/d/optout.
