Our SageMaker Studio service is broken in one of our AWS accounts in some deep way. Our original domain encountered this issue of "Update_Failed" when attempting to attach a new custom docker image. Using describe-domain via the CLI, we see that the "FailureReason" is just "InternalFailure". This issue also somehow effects brand new, entirely separate SageMaker Studio domains that we create. This is only an issue in our one (data science development) AWS account. Repeating the process in other accounts works as expected.
Are you using the exact image and app configs in other accounts? If so, can you please create a support case, since this would need a deeper dive into your settings and why domains in a single account is failing.
Yes, I've also tried attaching the conda-env custom studio image from this GitHub repo to rule out a bad configuration of the image or app config: https://github.com/aws-samples/sagemaker-studio-custom-image-samples/tree/main/examples
This also fails in exactly the same way in this one problem account. I opened a support case for this issue last week on Friday.