Skip to content

Conversation

@PatrickFranken
Copy link
Contributor

As the Kubernetes-based deployment of ERPNext requires RWX volumes, it's rather difficult to deploy on a cluster where the infrastructure provider does not offer such volumes. NFS is a workaround but also difficult to deploy right in a production environment. Due to this, IMHO it's a valid option to place all ERPNext pods on one single node which allows so share a usual RWO volume.
This scheduling can be achived using

nginx: 
  nodeSelector:
    system/ais-deployment: default
  affinity:
    podAffinity:
      requiredDuringSchedulingIgnoredDuringExecution:
        - labelSelector:
            matchLabels:
              app.kubernetes.io/app: frappe
          topologyKey: kubernetes.io/hostname

(The same has to be applied for all deployments, nginx is used as an example here)

This MR adds a common label to all pods, as this is required for this kind of affinity.

@PatrickFranken
Copy link
Contributor Author

@revant any chance to review and merge this? Not sure about the failed check, looks like it's not related to these changes. Let me know if you need any more information.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant