both algorithms are very low overhead and should not impose any noticeable load on the machine. could you let us know what type of machine you're running it on virtual or physical, how many CPU sockets and and whether you have hyperthreading, and also how many threads you're using for the backup. You'll find that last option in the Options dialog. And lastly when you say IO operations, do you mean CPU? Or something else?

Thanks for your reply.

No, it's not CPU - it's I/O operations as explained here https://www.oreilly.com/library/view/windows-server-2008/30000LTI00084/30000LTI00084_ch15sec22.html

This is the CPU in the last 28 days, obviously not loaded at all:



The machine in itself is not overloading because of these I/O operations. Just interested to know of anything that could cause this behaviour. You can see a spike in Disk Transfer/sec and read/write per second - the numbers are too small to impact the overall machine but higher than previous volumes.



Also note this machine has five drives but this change in I/O profile only happened on the boot drive - the other drives (SQL database, logs, paging file) aren't affected. It's a VM, 16 cores using 10 threads per backup plan.

In any case, I have asked support about this and will update later if anything comes to light.

Just additional information - that change upwards coincides with changing backup from Backblaze to Azure, enabling encryption/compression and Fast NTFS scan.

it may be related to the Fast NTFS scan option. You can try disabling it and see if the I/O returns to normal numbers.