Update 3a read entire disk

[jaseruk]

Hi all,

I installed update 3a on a customers server last night. A knock on effect of this seems to be that the entire disks were read, leading to the backups over running massively.

They have a Hyper V 2012 R2 cluster with up to date integration components. I have checked a customer with a VMware 6.0 cluster and I'm not seeing the same.

On a typical night we expect around 4.2TB to be processed, 100-150GB read and less than 30GB actually transferred.

Last night we had 4.3TB processed, 2.8TB read and then 29.5GB transferred.

We're mostly caught up now, just wondering if anyone else had seen anything similar?

There was no mention of CBT on the logs last night;

13/07/2018 00:07:47 :: Hard disk 1 (127.0 GB) 21.1 GB read at 32 MB/s

But on a follow up run I've done just now this is mentioned;

13/07/2018 10:53:26 :: Hard disk 1 (127.0 GB) 3.0 GB read at 30 MB/s [CBT]


I've had a dig through the release notes and forums posts but can't find anyone posting similar.

[Gostev]

Overall, it's not untypical for an update to reset CBT if it brings some fixes or enhancements to the Veeam CBT driver for Hyper-V.

[s90sml]

Ditto. Hyper-V backup that typically takes 2 hours ran for nearly 6 after applying Update 3a.

[nmdange]

One thing to note, this will only happen on 2012 R2, on Windows Server 2016, Veeam uses the new native RCT feature so it does not reset after an update.