aks questions - enterprise version
Has anyone used azure blob storage with code-server workspaces, three use cases
1. as an alternative to the default which is disk
2. as a shared read only for large files like the trivy dB in an air gapped environment
3. As shared read/write storage for large data manipulation tasks.
Coder is deployed in aks. I also have question about using deployment versus pods as I have seen templates for both. Does one have benefits over the other?
Also, are there any tricks to get envbox working in AKS?
If I use azure app registration as the oidc login or the external auth can I use that somehow to authenticate for the blob storage?
I do have workload identity enabled on the cluster but I am currently only using it for SOPS kubernetes secrets decryption.
I also plan to setup coder external provisioned. How can auto start and stop those so that I can scale the node pool to zero at night etc. ? Can I create a node pool with azure user assigned managed identity or do I have to use a service principal for the external provisioners?
Also if the external provisioners are in different clusters in different vnets in different subscriptions divided by different business units, do I have to use a multi cluster mesh with istio or is vnet peering sufficient.
1. as an alternative to the default which is disk
2. as a shared read only for large files like the trivy dB in an air gapped environment
3. As shared read/write storage for large data manipulation tasks.
Coder is deployed in aks. I also have question about using deployment versus pods as I have seen templates for both. Does one have benefits over the other?
Also, are there any tricks to get envbox working in AKS?
If I use azure app registration as the oidc login or the external auth can I use that somehow to authenticate for the blob storage?
I do have workload identity enabled on the cluster but I am currently only using it for SOPS kubernetes secrets decryption.
I also plan to setup coder external provisioned. How can auto start and stop those so that I can scale the node pool to zero at night etc. ? Can I create a node pool with azure user assigned managed identity or do I have to use a service principal for the external provisioners?
Also if the external provisioners are in different clusters in different vnets in different subscriptions divided by different business units, do I have to use a multi cluster mesh with istio or is vnet peering sufficient.
