Hello,

I tried google search and issue list to see if this is a reported issue.
Sorry if I have missed any of the discussion on the topic before.

We have an F1 instance with 1 QPS load. Our scheduler idle instance setting
is "Auto  - 3".

After about 5-6 hours of operation at that load, memory reported on the
dashboard goes up to ~220 MB. During this time, as expected, we see that
the scheduler spawns other instances. However, incoming  requests are
routed to the instance with peak memory. Then, understandably, we see
latency of the application really worsen ( because of memory constraints).
The instance hangs around for atleast 20-30 mins causing users to complain.
At some point the instance is killed and the requests are handled by new
instance(s).  If we manually kill the bloated instance switchover happens
immediately.

My question is :

Why does the instance hang around when the memory goes high, why isnt it
killed as soon as memory goes high. If the memory has gone beyond limits,
why the requests are being routed to this bloated instance when there are
other instances which have been created.

Thanks for any help.

Regards
Prashant

-- 
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?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.


Reply via email to