Hyper-V CSV Latency - Cumulative?

[JaySt]

I recently had a project with Veeam One v9 update 1 monitoring a Hyper-V 2012 (non R2) cluster. I wanted to see if i could show the disk latency throughout the cluster. The cluster is using iSCSI attached CSV volumes, arround 10 of them.
When i check the "Disk/CSV2012: Write Latency" (or Read Latency) counter at combined CSV Level, i get presented quite high latency values (30-50ms) on avarage. Drilling down to individual CSV volumes and looking at the same counter, values are much lower (<5ms). I almost seems like when i select the parent level of the CSV tree in the Monitor client, the client adds up all latency and shows a cumulative latency graph. I'm i looking at this wrong or....?
I basically could not explain why i'm seeing lower latency at the individual CSV level (all of them) compared to the parent level in the Monitor tree.

[Shestakov]

Hello,
Write latency on CSV shows the average latency between the time a write request arrived to the file system and the time when it was completed. Individual volumes should show the same for each volume and it`s not cumulative for the CSV. We will double-check if it works correctly.
How many volumes do you have? Do all of them show lower performance than the CSV itself?
Thanks!

[JaySt]

10 volumes in total. All volumes individually have lower latency values compared to the values graphed at the parent level.

[Shestakov]

Our QA team confirmed the bug. We are going to fix it in next releases.
Thank a lot for the heads up!

[JaySt]

ok no problem. Thanks for checking.
Would support be able to provide a patch perhaps ? if not, is there an ETA on the next update/release?

[Shestakov]

I will check with our QC team on a hotfix availability and will let you know. Thanks!

[Shestakov]

Could you contact our support team? Once the hotfix is ready they will provide it to you.
Thanks!

[JaySt]

ok i will. Thanks!

[Shestakov]

Once you do, please post your support case number for us to jump in just in case.
Thanks!