Disable SQL Warehouse during week-ends

[MaximeGendre]

Hello,
I massively deployed SQL Warehouses in our data Platform.
Right now, most of them are running every hour (with some inactivity phasis) because of Power BI report/jobs schedules.

To limit cost, I would like to stop/disable some on them on Friday evening and enable them on Monday morning.

Of course, I can ask people not to schedule things during the week end but with 500+ users it a losing battle.😣
The only way I can see would be to schedule some API call to drop all accesses, then reapply these accesses on Monday.
I see two drawbacks : 
     - It's not possible to delegate to trusted key users on project
            => they would need "Can Manage" and I don't trust them that much (I don't want to have XXL warehouses ...)
     - It would complexify access management on SQL Warehouse

Do you know if there is some 'disable' option / 'maintenance mode' that exists or which would be in the editor roadmap ?

Thank you

[szymon_dybczak · Accepted solution]

Hi @MaximeGendre,

So currently there's no easy way to disable sql warehouse entirely. Hence I think  approach you've suggested is valid. If I were you I would just create some groups and put your users into them. 
Then create a devops pipeline that will:

- remove all access groups from SQL Warehouse ACLs at the end of the Friday.
- re-add groups at then end of Sunday

If some users need to have higher privileges just place them into the group that will have i.e CAN MANAGE permission.

But yep, it would be great to have  'disable' option / 'maintenance mode'. Maybe we can create a ticket on databricks ideas portal regarding this feature?

[MaximeGendre]

Thank you @szymon_dybczak,
Let's go with API option then.
Can you please share with me the "best way" to submit an improvement idea ?

[szymon_dybczak]

Sure, here you have 2 options:

Feedback | Databricks

[nayan_wylde]

Also like to provide you with some alternate options.

• Tagging & Monitoring: Use tags and cost dashboards to monitor weekend usage and identify high-cost warehouses for manual intervention.
• Serverless SQL Warehouses: If not already in use, consider switching to serverless warehouses, which auto-scale and shut down during inactivity, potentially reducing idle costs