Reading speed per vm disk

[DanielJ]

I often observe that in VMware backups VBR reads a disk with a speed that looks capped or limited (a more or less straight line in the speed graph). We have more than enough bandwidth (and storage performance), but there is nothing to be gained from parallelism when a single disk (which might have several TB's of changed data) remains to finish in a job. Does anyone know of any such "speed limit per disk" in VMware, and if there is something that can be done about it?

[PetrM]

Hi Daniel,

For example, you may have a look at this topic but in your case we don't know what exactly affects read speed. What is the current speed and which transport mode is used?

Thanks!

[DanielJ]

Thanks, there were some interesting info in that thread. I wonder what became of the NFC re-architecture mentioned by VMware.

We use nbd with 10 Gbit interfaces, which is good enough except in these cases mentioned. We should probably start looking at hotadd proxy vms for this platform too. I never tested it in a vCD managed platform but as far as I understood it, there should be no difference from any other VMware platform.

[foggy]

Hotadd should just work with vCD, no issues there.