Clarify when certain disk operations entail a background copy of data

Question

Microsoft docs say the following about when the background copy process is invoked for disks:

The following control plane operations might move the disk between storage locations, and cause increased latency:

• Update the storage type.
• Detach and attach a disk from one VM to another.
• Create a managed disk from a VHD.
• Create a managed disk from a snapshot.
• Convert unmanaged disks to managed disks.

Can we predict when detaching and attaching a disk from one VM to another involves a background copy? I have a case where I want to avoid the performance penalty of the background copy. Does using VMs in the same region/availability zone guarantee that disks can be moved between VMs without the disk being copied in this way?

Answer 1

@Noah S Thanks for posting your query on Microsoft Q&A.

If both VMs are in the same region and availability zone, the likelihood of a background copy is reduced, but not entirely eliminated. The disk might still be moved between different storage locations within the same region, which can trigger a background copy.

Changing the storage type (e.g., from Standard HDD to Premium SSD) will likely involve a background copy, regardless of the region or availability zone.

For managed disks, Azure handles the storage location and type automatically. If the managed disk needs to be moved to a different storage location for any reason, a background copy might be triggered.