Lakehouse >Shortcut Delay

[Nicolas_Mattos]

Hello everyone,

I’d like to explain my situation:

We have a main Lakehouse that contains all our client’s data. We use Shortcuts to implement OLS (Object-Level Security). Each sector of our client’s company has its own workspace, and each workspace has a Lakehouse that contains only shortcut tables, restricting access to the data that users in that workspace are allowed to work with.

The main Lakehouse has a table called "date_time_2h", which is intended to record the last data refresh every two hours.

The problem is that some dashboards are refreshing with the wrong date_time. For example, a dashboard refreshes at 8:30 PM but displays the last refresh timestamp as 6:00 PM. All dashboards are scheduled to refresh 30 minutes after each pipeline run.

We've also noticed that some shortcuts cause more issues than others.

I understand that a slight delay between shortcuts and Lakehouses in Fabric is expected, but I believe a 30-minute delay shouldn't be happening.

Thanks for your help!

[v-hashadapu]

Hi @Nicolas_Mattos , Hope your issue is solved. If it is, please consider marking the answer 'Accept as solution', so others with similar issues may find it easily. If it isn't, please share the details.
Thank you.

[v-hashadapu]

Hi @Nicolas_Mattos , Hope your issue is solved. If it is, please consider marking the answer 'Accept as solution', so others with similar issues may find it easily. If it isn't, please share the details.
Thank you.

[v-hashadapu]

Hi @Nicolas_Mattos , Hope your issue is solved. If it is, please consider marking the answer 'Accept as solution', so others with similar issues may find it easily. If it isn't, please share the details.
Thank you.

Thank you for your prompt response @nilendraFabric .

[nilendraFabric · Accepted solution]

Hi @Nicolas_Mattos 

Here is main reason of delay :

Shortcuts rely on periodic metadata synchronization (typically in few minutes), which combines with pipeline execution and dashboard refresh schedules to create compounding delays. The `date_time_2h` table’s updates must propagate through multiple synchronization checkpoints:

Main Lakehouse → OneLake Cache → Shortcut Metadata → Dashboard Query

Your OLS implementation using cross-workspace shortcuts introduces additional validation layers at each workspace boundary. Each security checkpoint adds 2-3 minutes to data propagation

OneLake maintains cached snapshots of Warehouse data for query performance, with refresh cycles that prioritize stability over immediacy. Critical timestamp tables like `date_time_2h` become particularly susceptible to this latency.

you can try doing this , lets say schedule every 5 mins or whatever feels right in ur scenario;

from semantic_link import refresh_metadata
refresh_metadata(lakehouse="main_lh", tables=["date_time_2h"])

see if this helps

thanks