You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
TF provider assume there will always be 1 and only 1 public network within a resource instance.
As observed on UI the public network is created automatically when a VM with public network is created. Also once the allocation pool is completely utilized it will create a new public network with different subnet. Still, TF provider will return the first indexed network.
The text was updated successfully, but these errors were encountered:
The workaround here is to always create a new public network and use that for the bastion node.
In such scenario, a new public subnet will be created for every cluster on the resource. This can add to the cost but needed to have stable deployments.
We also get the public network gateway address if we follow above approach. Which is currently not possible using ibm_pi_public_network data source.
Potential Issue: TF will fail to delete the public network when a cluster is destroyed. This will happen when other machines (manually created from UI) use this public network. This can be mentioned in the known issues.
The terraform module will crash with index out of bound error as mentioned in IBM-Cloud/terraform-provider-ibm#1701
TF provider assume there will always be 1 and only 1 public network within a resource instance.
As observed on UI the public network is created automatically when a VM with public network is created. Also once the allocation pool is completely utilized it will create a new public network with different subnet. Still, TF provider will return the first indexed network.
The text was updated successfully, but these errors were encountered: