0

I've been trying to deploy an app onto appengine, and i keep getting this error while trying to load the webpage. The deployment has gone well with no erros.

I've seen a very similar question about this issue and tried to increase the RAM to max (B8-instnace).

I will say the app is high-mem demanding, but it runs locally.

my app.yaml file

runtime: python37

instance_class: B8
basic_scaling:
  max_instances: 11
  idle_timeout: 10m

Is it possible it's not a memory issue and actually something else? Any help with the issue will be great.

oriperl
  • 1
  • 1
  • 2
  • hey @oriperl , the issue here is indeed memory related as your application is comsuming memory(beacause of libraries like plotly-dash) more than what the instances in your appication can handle and fails with the error that you posted. to help you, can you confirm few stuff: is your application working but receiving that error or it's not working at all? can you post your app.yaml? – Methkal Khalawi Dec 31 '19 at 09:23
  • hi @MethkalKhalawi, i edited my yaml file into the question. And about the app, it is not working. thanks! – oriperl Jan 01 '20 at 07:47

2 Answers2

0

In your case B8 machine type 2GB memory is not enough.

Firstly, you need to try to reduce application memory usage. Review your application code and dependencies in case you can decrease memory consumption.

If you cannot decrease your application memory usage, you can use App Engine Flexible environment and set up Resource settings with more that 2GB memory in.

Also you can create an instance with appropriate settings on Google Compute Engine and deploy app on it.

Jaroslav
  • 724
  • 4
  • 17
  • Hey, thanks for the reply! im taking your advice and trying to deploy the app on the comute engine. thanks! – oriperl Jan 13 '20 at 09:44
0

This is not a message indicating you have run out. It is merely a warning from OpenBLAS, a numerical library, that it could not determine you cache size and therefore perform particular optimizations. You might be loading the library with Numpy or Pandas or similar. I haven't yet found a way to eliminate, but it can just be ignored.

Kevin Buchs
  • 2,520
  • 4
  • 36
  • 55