Google Cloud Datalab not running cells: Folder Creation and Save and Checkpoint Failures

I am finding that I can't save, run or otherwise interact with Google Cloud Datalab as of today at about 2pm AEST (Australia Sydney Time). The datalab is run in the region of Australia-1a and has a size n1-standard-1. I get the error "Creating Folder Failed: An error occurred while creating a new folder" and saving just doesn't work.

An attempt to 'Save and checkpoint' results in "Autosave Failed!". I suspect this may be because of Datalab updates but when I run:

gcloud components update

gcloud components install datalab

I get that all components are up to date. I can still view previously created .ipynb files and start/stop compute engine instances but both for pre-existing and new datalab instances I get the same errors and cannot run any code. Attempts to run code cells do not give any error message but rather just do nothing.

As an update, I was able to use Datalab on a vm located in Australia (like the above example) if I switched to an OLD project. On my previous project If I switch to a different project I am able to use Datalab but I get the notification that:

"You are using DataLab 1.2.20180511. An optional update (1.2.20180608) is available (see whats new)."

Switching back to my original project I do not get that notification, but I still cannot run cells or save folders. If I create a brand new project likewise, all the same errors occur. Only my very old project still works.

Is it possible that somehow the update has been applied and is causing the issue? If so, is there anything I can do to roll it back?

We published a new release yesterday that was meant to improve the security of Datalab instances by rejecting cross-origin requests.

However, it looks like that release prevents Datalab connections from Cloud Shell from working.

Sorry that you hit this; I am working on rolling back the release now.

EDIT: The rollback has finished and the problem should be resolved if you recreate your instance. We created this issue to track finding and fixing the root cause of the bug.