According to the company, bursting lets eligible disks achieve up to 30x the provisioned performance target to better handle spiky workloads. This should help in workloads that have unpredictable disk usage and improve overall performance. Microsoft warned in its documentation and supporting blog post that this functionality will only be available on its premium SSDs. It will also work on a “best effort” credit-based system. Essentially, you get credits when the traffic is below your provisioned target, which are consumed when it exceeds that through burst. As well as spiky workloads, Microsoft says this will help with the virtual machine boots. During startup, the OS disk may be used at a higher rate than what’s provisioned, and burst could also reduce application launch time. “With this preview release, we lower the entry cost of cloud adoption with smaller disk sizes and make our disk offerings more performant leveraging burst support,” said Azure Storage program manager Yuemin Lu in the blog post. Those smaller burst capable disks are P1, P2, and P3, bringing the total number of burst capable disk types to eight. All support max bursts of up to 170MiB for up to 30 minutes at peak. Burst IOPS goes up to 3,500 on each disk. Here are the full stats: Unfortunately, you’ll only be able to use this function currently in the Azure West Central US region. Microsoft has plans for a wider rollout soon, though, and the good news is that it’ll be enabled by default in all regions.
