Introducing Surge Protection (Preview) · Learn Microsoft Fabric

🚀

Introducing Surge Protection (Preview)

Last week, Microsoft introduced a new feature: Surge Protection (preview). You can read the full article here if you want more info, and also the related documentation page for Surge Protection.

𝗦𝗼 𝘄𝗵𝗮𝘁 𝗶𝘀 𝗦𝘂𝗿𝗴𝗲 𝗣𝗿𝗼𝘁𝗲𝗰𝘁𝗶𝗼𝗻?

Essentially, surge protection prevents background jobs taking up too much of your capacity allowance.

A background job is typically a scheduled job, like a scheduled pipeline run, or a scheduled semantic model refresh. As opposed to interactive queries - these normally involve a user clicking on buttons in the UI. Both interactive and background job capacity usage are reported in the Capacity Metrics App.

𝗛𝗼𝘄 𝗱𝗼𝗲𝘀 𝗶𝘁 𝘄𝗼𝗿𝗸?

You define a threshold ("Background Rejection Threshold"), for example 75%. This means that when background jobs are consuming >75% of your capacity, they will begin to be rejected.
You define a second Recovery threshold, which could be 50%. When the Background Recovery threshold is reached, surge protection stops rejecting operations. The capacity then resumes accepting new background operations.

𝗜𝗺𝗽𝗹𝗲𝗺𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻 𝗻𝗼𝘁𝗲𝘀:

You configure surge protection for each capacity.
Before you set the two thresholds, check your capacity’s 24 hours background %. Avoid setting the value too low too quickly to prevent impacting users.

Let me know if you'll be turning on this feature! Please note: this is still a preview feature, so you'll need to bear that in mind.

2 comments