Saving and publishing data flows
As you work on data flows, you are likely to make edits to the definition, such as creating new actions or connections. When you edit the data flow, you create a draft. If your data flow is shared with other users, they will also be able to edit the definition, creating their own drafts.
In addition to creating and saving drafts, you can publish a draft, making these changes available to users with whom the data flow is shared.
When you initially create a data flow, it is automatically published.
Your data flows can be in one of several states as indicated in the header bar next to the data flow name:
-
Draft Unsaved - The draft is currently unsaved, typically with pending edits.
-
Draft Saved - The current draft has been saved (there are no pending edits).
-
Published - The current draft has been published and is available to other users.
If you attempt to close a data flow that has pending changes, you are asked if you want to save or discard the changes.
-
Save - The changes in the current draft are saved and the data flow closes.
-
Discard - Current changes are discarded and the data flow closes.
-
X (upper right) - The confirmation message box is closed. The data flow remains open with pending changes.
Update data flow properties
-
Click the kebab
on the header bar and select Data Flow Properties. The Data Flow Properties pop-up is displayed.
-
Edit the Name and Description as needed. The Description is limited to 256 characters.
-
Click Save.
Drafts and publishing
As you edit your data flow definition, a draft is created. This draft is initially in an unsaved state (shown as "Draft" in the header bar). If you close the data flow when you have pending edits in the current draft (that is, an unsaved draft), you are prompted to save (or close without saving changes).
Please note the following:
-
You can save a draft even if there are validation errors. This allows you to save your work so far and continue to edit the draft contents.
-
You cannot publish your current draft if there are specific validation errors. For example, if you have invalid nodes that are part of a flow that originates with the Start node, the draft cannot be published. However, if there are invalid notes that do not have an origin with Start (an "island node"), the draft can be published.
The header bar provides features to manage your drafts:
-
You can save drafts at any time. When you click Save in the header bar, you will see "Draft Saved" as the data flow state.
-
You can make your current draft available to users with whom the data flow is shared. When you click Publish, your current draft is validated to ensure that required fields are populated and there are no validation errors. If the draft is valid, it is saved and made available for all users of the data flow. You see "Published" as the data flow state.
If you have never successfully run the data flow, a warning message is displayed when you click Publish. It is recommended that you run the data flow before publishing to be sure that it is configured to run successfully.
-
If you decide you do not want to keep the current draft (saved or not saved), click the kebab
on the header bar and select Revert to Published. You are asked to confirm this action. Click Revert. The current draft is discarded and the data flow reverts to the last published draft.
-
If you decide you do not want to keep the current unsaved changes to your draft, click the kebab
on the header bar and select Discard Unsaved Changes. You are asked to confirm this action. Click Discard. The unsaved changes are discarded and the data flow remains in Draft mode.
Drafts and multiple users
When you open a data flow, you will see the last published draft or your own last saved draft. Keep in mind that this may be one published by another user with whom the data flow is shared. When you are working on a data flow and have a draft (saved or unsaved), this version of the data flow is currently only available to you. You cannot share a draft with another user. Other users with access to the data flow will see either the last published draft or their own current draft if they have edited the data flow.
For example, assume User A and User B are going to work on a shared data flow:
Action | State when opening the data flow |
---|---|
User A creates the data flow. | Data flow is published. |
User A shares the data flow with User B. | Both User A and User B see the published draft when opening the data flow. |
User A creates draft 1A. |
User A sees draft 1A. User B sees the original published draft. |
User A publishes draft 1A. |
User A sees published draft 1A. User B sees published draft 1A. |
User A creates draft 2A. |
User A sees draft 2A. User B sees published draft 1A. |
User B creates and publishes draft 1B |
User A sees published draft 1B. User B sees published draft 1B. |
Draft validation
Invalid conditions in the draft, such as duplicate action names or incomplete action definitions, are identified with warning icons (red exclamation marks) and validation text. You cannot publish a draft that has invalid conditions.
Last modified: Wednesday September 11, 2024