JunkDoom

@GrantSmith Thanks! After some trial and error and confusing schema warnings, here's what's working for me so far: -Create a new REPLACE job, and RUN it. -Change the job to APPEND, which will then allow you to pick an UPSERT key in the "schema" tab. (may need to uncheck "Protect" on that column. -Change the query to look…

The lack of documentation on Partitioning in Workbench gives me pause. If we're not specifying an UPSERT key, wouldn't this only work for transactional data that never changes at the source? For example, say I create a "Replace" job for an initial load. I partition on a "DateUpdated" field. I then change the query to only…

I am able to create multiple views of the same dataset, such as "View_Sellers" and "View Buyers", then join that to Transactions. Only issue there is that Domo doesn't distinguish which field is from which table. It just appends a number to the column name automatically. For example, "First Name", "Last Name", Contact ID",…

Perfect! Makes sense to me. I converted a couple, and things seem to be working so far! Thank you.

A-ha! I see it there. I also see that DataSet views look to be counting against our row count, which was not my understanding at all. I will contact support to see what counts and what doesn't since it's not clear anywhere. Thanks!

@grace@GrantSmith Even if it is private, which is does appear to be, it's still concerning. Hopefully Domo will change this in the future so there's no confusion.

Not ideal, but if it's not just happening to me, I'll count on sharing everything as needed. Thanks!

Grant - great point on using a custom role for safety here! I ended up submitting a support ticket for this, and so far, they think the behavior I'm seeing is a bug.

Great! This is exactly what I need. However, there is no mention of how row count affects licensing. Do row counts in views count against our licensing or are these 'free' since they're just a join of existing data sets?

Thanks. That makes sense. However, it's a little concerning that Workbench doesn't let you change the dataset type on existing jobs. Perhaps I'll just use SQL Server type going forward, and leave the existing ones as is.

I found something close to what I was looking for. In Magic ETL in the left panel, grab a "Replace Text" function (In the "Edit Data" section). Drop this into the data flow. In that function select a column, and in the "Enter a term to search for" box, click the gear to change the search method to "RegEx". Then use the…