* Mount delete, refresh and default to new folder for mount create
- Delete mount action added
- Refresh mount action added
- Added "mymount" to theend of existing path so that we don't use already-existing HDFS folder. The call fails unless folder doesn't exist
Implemented in this PR
- New base dialog for anything needing to work with the controller. This is important since going from SQL -> Controller we "should" have the right permissions but aren't guaranteed
- Support for Mount HDFS via a dialog. Includes basic polling for success/failure, but have to give up after 2.5min as mounting could take hours. By then it's assumed to be successful since server-side has 2min timeout built in.
Not implemented in this PR
- Script as Notebook button. This should convert the inputs to a set of cells in a notebook so users can run things themselves
- Updates based on PM / UX reviews. I think we'll need a round of feedback before completing this work.
- Use token API to get auth token before using other endpoints. Note this needs server updates before it'll work as expected (will only checkin after verifying this)
- Add auth option in controller UI and plumb through connection save, load, and uses of the controller API
Because the swagger spec is split in 2 created new file for the 2nd swagger spec including token auth endpoints. These come from a running cluster and instructions were updated to reflect this.
New UI Changes:
- Added authentication type field with "Basic" and "Windows Authentication" as the options
- Moved error notifications to the dialog instead of separate notification window. That's the recommended pattern
- Username / password aren't required for Windows Authentication. I couldn't find a way to change required status in form container on switching from Windows => Basic Auth so have error show on clicking OK.
- Controller URL should use ":" not "," for IP:Port
* Fix#6477 controller login + fix dashboard layout
- Service endpoints shoudl be on own column, cut off smaller screen
- Controller login not working due to 404 error
This is due to a breaking API change
We have requested fixes to help mitigate need for cluster name,
but for now have a default value for this
Finally, modified code so it's easier to update swagger API
and also added instructions on how to update in future
