Deconfusing the image creation process.
What I heard in the meeting today:
-
Replace Clone/Snapshot buttons with a single Create Image button.
-
We show this button all the time.
-
When the user clicks on the button:
-
If the user is allowed to Clone or Snapshot, we present them with a choice; Do you want to create a new profile or update the current profile. Clearly if the user is not allowed to update the current profile, then we do not need to show this choice, new profile is the only choice and we skip this step.
-
If creating a new profile, then we ask the user if they want to start with an empty profile or make a copy of the current profile.
-
If there are multiple nodes in the experiment, we ask the user to select via dropdown, which node they want to snapshot. Aside; we should probably remove the Snapshot button from the node context menu and the List view to reduce UI clutter/confusion.
-
Ask the current set of questions with checkboxes:
- Update the profile to use the new disk image?
- Did you add any accounts (for prepare)?
-
Do the imaging, show the imaging progress modal.
-
If the operation is a plain snapshot, update the rspec if requested and its not a geni-lib script. We are done.
-
If a new profile has been requested (clone), take the user to the new create profile page, either with the updated source code (if not a geni-lib script) or with an empty profile. Either way, need to tell the user the URN of their new image (and pass that to Jacks for people using the topology editor).
-