I have noticed that when deploying with a subdomain name after cleaning a previous deployment with the same subdomain name, the AWS keypair gets borked.
During cleaning, I have verified the the AWS keypair resource and the local copy of that keypair are both deleted.
On the new run, I have verified that a new keypair is made in AWS and that new key is propogated to the local host.
After completion, I have verified that the new keypair does not match and I cannot SSH to the node. I have also verified that if I save the old keypair from the previous run off to another location, clean bad deployment up, deploy new then SSH to the instance with the old keypair, I can connect just fine.
Appears to be a bug with AWS but unlikely since they are AWS and I am a shitty bash scripter.
I have noticed that when deploying with a subdomain name after cleaning a previous deployment with the same subdomain name, the AWS keypair gets borked.
During cleaning, I have verified the the AWS keypair resource and the local copy of that keypair are both deleted.
On the new run, I have verified that a new keypair is made in AWS and that new key is propogated to the local host.
After completion, I have verified that the new keypair does not match and I cannot SSH to the node. I have also verified that if I save the old keypair from the previous run off to another location, clean bad deployment up, deploy new then SSH to the instance with the old keypair, I can connect just fine.
Appears to be a bug with AWS but unlikely since they are AWS and I am a shitty bash scripter.