Advice on multi-lookup column list threshold workaround

Copper Contributor

Hello,

 

Situation:

We are tagging most our data in Sharepoint with project numbers via a multi-lookup column.

Now some lists start to reach the threshold limit, and you can't index multi-lookup columns.

 

Problem:

Losing the functionality to filter, aggregate or sum by the project number is not an option. Neither is using a single-lookup column nor keeping the lists small.

 

Possible Solution?:

Are there any known best practices, so that the user can still filter and our pnp-js and pnp-powershell code can make use of an indexed column? By archiving the lists and libraries will cap at around 10 000 - 50 000 items.

 

I was looking at making use of the term store, a workflow to automatically mirror the project numbers to the new column and a powershell script to update the term store.

 

But before doing this, I wanted to ask more experienced users, if this makes sense and if there might be other solutions.

 

Best wishes

Nico

 

4 Replies

@de_simon I have the same issue.

 

by pnp-js are you referring to the web parts?

 

I have the added issue of our powerapps that access project numbers and 3 other mmd (managed-metadata) fields "becoming unpredictable" = "very bad" when the list view exceeds 5000.

 

by removing all other fields, these indexed columns do work for us on most views, the line is thin though, if I add one more column such as "modified by" I get an error.

 

Yes, we are using a web part to create user forms. It allows me to run js.
Hello,
Did you ever figured out a solution for this problem of yours. I'm having the same issue and not sure how can I use PnP JS to get around this. Please let me know if you found any possible solution.
Thanks,
Suhas
Hi,
no there is no real solution. What I'm doing is using pnp-powershell scripts to aggregate and sum the data on a local server and write it back into a sepprate sp list.

Also we are using plumsail forms, this replaces the standard forms. There i can run javascript and recreate certain functionalities with pnp-js.

In the long term we are planing to move the data to a SQL Server and use PowerBI to visualize it.