Hi, I'm new to flyte and kubernetes. I followed the flyte the hard way tutorial, and was able to get the deployment up, with an ingress through an ALB. However, when I try to run the hello_world example, a pod is created, but then fails to proceed further. Here's the event output for `kubectl describe pod <podname> -n flytesnacks-development`:

Events:
Type     Reason            Age   From               Message
  ----     ------            ----  ----               -------
  Warning  FailedScheduling  67s   default-scheduler  0/4 nodes are available: 4 Insufficient cpu. preemption: 0/4 nodes are available: 4 No preemption victims found for incoming pod.

What would be your recommendation course of action for de-bugging this problem?

what instance type are you using? also, you can reduce the default CPU requests from the Helm values file:

task_resources:
      defaults:
        cpu: 500m

I'm glad the tutorial got you there, and I'm really interested in learning from this to adjust the values to work OOB

Thanks David. Instance type that is currently being created are t3.medium. I'll try reducing the default CPU requests. It looks like I'm not specifying

task_resources

, so will add that to the Helm values file.

cc @David Espejo (he/him) can you help here? It seems like the nodes are not autoscaling or the node configurations is wrong, that we do not have nodes that can have 100Mi cpu?

The pod name is

a24gnswfz78p92ggx8jh-fvub2h2y-0

is that what you're asking?

@Ketan (kumare3) I'm talking to Ryo, while there's no documented resource consumption profiling for Flyte AFAIK, I've seen

t3.medium

(2vCPU) not being enough in the past. We'll test with

t3.xlarge

Following up here. The switch to

t3.xlarge

worked. Thank you @David Espejo (he/him) for your help.