0/1 nodes are available: 1 Insufficient cpu. preemption: 0/1 nodes are available: 1 No preemption victims found for incoming pod.. (x44)

[moisesrodriguez]

INFORMATION
Cluster

ISSUE

When deploying it says 0/1 nodes are available: 1 Insufficient cpu. preemption: 0/1 nodes are available: 1 No preemption victims found for incoming pod.. (x44).

I have a database with the following resources allocated:

And the app that I want to deploy with these resources allocated:

My cluster has 2vCPU and 4.0 GiB of memory. So if I do the math of the database with .5vCPU and 512 MiB plus the same for the app, it gives me a total of 1 vCPU and 1024 MiB of memory. Still under what the cluster has available. What am I missing?

HOW TO REPRODUCE

Deploy application

[moisesrodriguez]

Just for the sake of it, I upgraded de cluster to a t3a.large cluster

and still, I’m having the same issue

[Erebe (Romain Gerard)]

Hello,

Would you mind sharing the console.qovery.com url of your environment so I can take a look at it ?

[moisesrodriguez]

@Erebe
console.qovery.com

[moisesrodriguez]

@Erebe

I switched to an EKS cluster and with the same settings it deployed.

[Melvin_Zottola]

Hello @moisesrodriguez ,

You had this issue because the resources were too low,
you can read this answer that explains the resources details on k3s: CPU and RAM Usage on EC2 K3S instance with Qovery - #4 by rophilogene

[moisesrodriguez]

The thing is @Melvin_Zottola is that I upgraded it to t3a.large and still had the issue. It only went away when switching to a eks cluster

[Melvin_Zottola]

Hello @moisesrodriguez,

As stated in your deployment logs, it was due to missing vCPU resources:

Unfortunately, t3a.large has the same vCPU spec than t3a.medium (it has just more memory)

[moisesrodriguez]

@Melvin_Zottola

For this one case don’t know what the issue was, I deleted everything (cluster included), and 2nd time it worked with the same specs and app resources. It can always be the case I did something different, but on paper, it was the same.