I am a subscriber of the Always On option in Google AppEngine. I don’t understand why GAE should start new instances when there are Always On instances already included, and that it does not have a large number of requests at that time.
This drives me crazy, as it negates what I get with an AlwaysOn subscription. Any ideas how to fix this behavior? I have a delay problem due to this that I wanted to remove with AlwaysOn. I am thinking about canceling the subscription as it just doesn’t work.
, AlwaysOn, , - , - . , , , , App Engine .
, GAE (.. , ). , , , .. (, , ) GAE .
, , , AlwaysOn . , , GAE - .
, Always On, , -, , , .
Warmup Requests, App Engine , , , , .
, , . , ? , , . .
I just wanted to add the answer I received on an identical question, which I posted in the google-app-engine group ( Link ). This has helped me reduce the likelihood that new nodes will be created when some active nodes exist at that time.
Do you have "threadsafe" set to true? If you do not, each instance will process one request at a time.
It helped me a lot - try it if you have the same problem as me. Using_Concurrent_Requests