We have a Hyper-V Cluster through SCVMM 2012 SP1
All hosts are Server 2012 - All are HP DL360p Gen8, 2x E5-2650 (32 Logical Cores), 256GB of memory - All are connected to 3x4GB Luns on an EMC vnx5300
Our backup host is a DL180 G6, 1x Xeon E5620 with 24GB of Memory, running server 2008 R2 SP1 and Veeam 220.127.116.11 (64Bit).
During our Backup schedule using Veeam we get (what appears to be intermittent) failures of applications, including our web servers, telephony control services and even delayed mail.
I have run some monitoring of the performance collection stats and I can see that the CPU spikes drastically during this period vs normal operation.
Our CPU contention ratio is about 1:0.73 (1 logical core per 0.73 vCPU assigned). 96 Logical cores with 132 vCPU assigned.
Network throttling still doesn't alleviate the issue (at all).
Your help is appreciated.