1

I have several services which have very little load (several requests per week sometimes). During periods of constant use there appears to be no problems. But after a period of being idle instances appear to be spun down. Web console shows that instance is “being restarted”. In my app.yaml I have automatic_scaling with min_num_instances: 1. However instances continue to be “spun down”. I tried adding min_idle_instances: 1 as well, however, I got an error (which makes sense):

VM-based automatic scaling should NOT have the following parameter(s): [min_idle_instances]

Question is how do I prevent the VMs from becoming unavailable/restarting? Am I understanding correctly that scaling down is the cause? Or if it’s not, where can I see some information about the reasons? Instance’s logs look normal to me. Dashboard, Error Reporting, etc. also look rather innocuous. Again, when instance receives requests (or after being re-deployed) it behaves normally without any noticeable problems.

beyondcompute
  • 11
  • 2
  • Did you ever find the solution to this? I am having similar problems – Nicholas Nov 01 '18 at 03:36
  • Hello, @Nicholas! No, unfortunately I still haven’t found out what causes the problem. – beyondcompute Nov 02 '18 at 11:32
  • Is the entire application becoming and staying unavailable after periods of low load, or are there shorter periods of unavailability, or are you asking about the behavior of the individual VMs? – Jesse Scherer Dec 17 '18 at 22:54

0 Answers0