Latency metrics for disks and performance metrics for temporary disks on Azure Virtual Machines

We are excited to introduce the capability to monitor latency (in preview) across OS, data and temporary disks using the SCSI protocol (default), with support for the NVMe protocol coming soon. This enhancement is particularly important for high-performance applications such as SAP Hana and OLTP databases, where latency plays a crucial role in read and write operations. Now, you can track the latency of your disk operations on OS, data, and temporary disks using Azure Monitor metrics.

In addition, temporary disks, which are inherently non-persistent and are typically found in Virtual Machines (VM) families marked with a 'd' in their name, such as the edsv5 family, have now become accessible for performance monitoring on Azure Monitor. You can now easily track queue depth, IOPS and throughput for these temporary disks, even though their storage does not endure beyond the lifecycle of the associated VM.

To learn more, read the documentation