Add UX improvements to indicate why creating a pipeline after uploading a file in data prep does not add a source
Description
Release Notes
Attachments
Activity

Bhooshan Mogal September 11, 2018 at 9:25 PM
Deprioritizing because of time constraints. Will take this up in 5.2.0

Bhooshan Mogal September 8, 2018 at 12:34 AM
Moving to next sprint for implementation.
Lea Cuniberti-Duran September 7, 2018 at 9:24 PMEdited
Finalized user flow:
The user selects a file from "upload"
After applying some transformation, the user selects the call to action "Create a Pipeline".
Modal opens to prompt the user to select the type of pipeline (batch or realtime). In the modal a warning is displayed.
Please apply same visual rendering than the warning messages in dataprep. (see attached image)
In pipeline studio, only the wrangler node is displayed.
*Modal text:*
Warning
You are creating a pipeline using a file that has been uploaded from a local source. Since the source cannot be inferred, you have to manually configure a source for the pipeline.
Acceptance Criteria
Add a warning to the pipeline type modal
Display only wrangler node in studio

Bhooshan Mogal August 29, 2018 at 6:45 PM
could you please help drive this decision?

Sreevatsan Raman August 17, 2018 at 3:46 PM
Do not vote to remove. Upload is a useful functionality in dataprep. The Sandbox is available in several environments not only for laptop. We have Cloud sandbox in Amazon and Azure. The upload should be supported in those cases.
It would be good for users to know what do to before taking an action, so i recommend option 4
Details
Details
Assignee
Reporter

The difference between FileSystem and Upload connections is not apparent when you are running a CDAP sandbox on your laptop. It is much clearer when you are running on a cluster, in which case, Upload reads from your laptop, where FileSystem reads from the cluster’s HDFS.
It would be helpful to add some hints to reduce ambiguity/confusion.