Schedules getting Triggered every hour after moving to cloud.

[rsareen23]

We used SDK to migrate the contents from on prem to cloud. Dashboards/Datasources migrated with no issues. However we see that after migrating, schedules are are running for every one hour. However I see "Every Mon,Tue,Wed,Thu,Fri at 7:00 AM(UTC-05:00) America/newyork" on the dashboard/data sources on the "extract refresh" tab.
If we create new schedule manually, it is working as expected. Since we have thousands of schedules, creating schedules manually is not feasible.
Any Thoughts/inputs?

[pvshn]

What does the schedule look like on Server ? Only a few schedules are allowed on Cloud. So, there is conversion in the Migration SDK code. I want to make sure the translation is happening correctly.

[rsareen23]

Onprem looks like- All days of the week, at 9:00 AM (UTC-05:00) America/New_York (next run at: Oct 1, 2025, 9:00 AM)
This schedule on cloud looks like - All days of the week, at 9:00 AM (UTC-05:00) America/New_York (next update: Sep 30, 2025, 2:00 PM)
basically, cloud one running every hour.
Note: we are using the Python based SDK for Tableau migration

[michaelhrichardson]

We saw this behaviour on both the CommBank migration and on the Westpac bank migration. In the first instance the customer wrote a Python script to download all the schedules, delete them and then recreate them. It was a little tricky as the schedules populate different fields depending on whether the schedules is hourly, daily, weekly or monthly. I don't have a fully working version of this script, unfortunately.

Westpac had fewer instances. We took a manual approach to fix this (and also to adjust each schedule so it didn't trigger on top of the hour - to reduce concurrency pressures on the Bridge clients). I created a workbook on top of the Jobs data source in Admin Insights. It listed each workbook/data source with extract refresh jobs, and visually displayed each instance of the extract refresh. This clearly showed the 'rogue' jobs as they rendered as tight clusters marks on the viz. Using a URL Action we were able to go straight to the workbook/ds and edit the schedule. Simply resaving the schedule seemed to fix it. When admin insights refreshed the following day the number of marks significantly reduced.

[rsareen23]

Understand, Thank you for clarification. we will proceed manaully fixing the schedules for now.

[pvshn]

Are all weekly schedules being converted to hourly or just this one?

[pvshn]

Could you post the server and cloud schedule with the intervals and frequencies along with the start time you mentioned earlier?
I want to check if it is really a conversion issue or something to do with the REST API.

[michaelhrichardson]

Hi Pranav. Both the CBA and Westpac Bank migrations are now completed, and I have since been offboarded from both banks, so no longer have access to their Tableau Cloud environments. This issue was significant for both customers, and led to a several days effort to manually readjust these schedule.

[michaelhrichardson]

Are all weekly schedules being converted to hourly or just this one?

We found that this issue affected daily schedule refreshes. They were converted to hourly refreshing when the schedules were migrated with the Migration SDK.

[pvshn]

We have a Migration SDK side fix we expect to come out with the next release. A Tableau REST API fix is also needed. However, the Migration SDK fix is going to work around the fact that the default hour interval being 1.