Save AS ETL not functioning properly?

HowDoIDomo

Hi, I am running into a weird issue.

In most cases you can use "Save As" to make adjusted copies of a card or dashboard without changing the original. However, that's not working for ETLs for me.

Instead of having a separate ETL, it keeps sending me back into the original. I'm just trying to apply the same name mapping as the original, but to a slightly different dataset. If I copy and paste the tiles into a new ETL, I would still have to re-enter all of the mappings.

Does anyone have an idea on how I can expedite setting up an ETL when I already have the logic and mappings in another?

Thank you!

MichelleH

@HowDoIDomo It usually works for me if I just the "Create Copy" option

HowDoIDomo

@MichelleH Thank you! I'll try that. I'm realizing now that If I change the input file, I lose all of data entered into any subsequent tiles. is there any way to change the sourcefile without losing my previous work?

MichelleH

@HowDoIDomo Are there specific tiles where you're losing work? It may help to see a screenshot of what you're seeing. Also, when you change the input file are you adding an entire new Input tile or just changing the dataset on the same tile?

ArborRose

Domo datasets have a unique identifier behind them. When you view a dataset, you will see the identifier in the URL. When you make a copy of an ETL, you are normally setting up new datasets rather than pointing back at the original source of the tile. This pertains specifically to output tiles vs inputs.

Output tiles default to replacing the dataset rather than appending the dataset.

• Unique Identifier for Datasets:
  • Domo datasets have a unique identifier visible in the URL. This identifier is what ties various elements (tiles, dashboards, etc.) to the specific dataset.

• Copying ETLs and Dataset Disconnect:
  • When you copy an ETL, it usually creates new datasets for the output tiles rather than pointing back to the original dataset. This is crucial when you're duplicating ETLs for a new use case, as the new datasets act independently of the originals.

• Output Tiles vs. Input Tiles:
  • Output tiles specifically create or update datasets in Domo. By default, these output tiles are set to replace the dataset on each run, not append to it. This default behavior is important to highlight because it ensures the dataset reflects the current state of the ETL logic without accumulating historical data unless specifically configured to append.

Ashleigh

@HowDoIDomo I just discovered Save as in ETL today. Using Save As will make a new flow but its still connected the the same output dataset. This makes the dataset ID no longer unique. Not its one dataset being powered by 2 flows (or more) and now either flow can make edits to that dataset.

To get a complete new copy where you just want to copy logic and have it in a new dataset you will need to do what @MichelleH mentioned and use Create a Copy.

I have not been able to find any documentation about how Save As works but just know it will create new flows that all power the same output dataset, hence why it keeps bringing you back to the same one, because it is the same one.