suggestion for improvement?

[SpiderPig]

When the backup begins all the vms on the host slows to a crawl and its pretty much unusable, slow and sluggish.
We use onhost-proxy for the two jobs (ca 55 vms pr job). The Hyper-V cluster show no sign of distress during the backup, 4-5% cpu, the vnx shows a little more disk io then usual and the Emc Datadomain with ddboost target have no issues.

Will the use of an off-host proxy do any good in this scenario?

A typical day results in this:

Job 1 (58 vms)
Duration 8:14:10
Procrate: 32MB/s
Bottlnck: Source
Processed: 6,8TB
Read: 688,4GB
Transfrd: 399,4GB

Job 2 (53 vms)
Duration 5:58:36
Procrate: 17MB/s
Bottlnck: Source
Processed: 2,1TB
Read: 137,2GB
Transfrd: 86,9GB

[Shestakov]

Hello,
Yes, off-host proxy is to shift the load from the host during backup job activity.
Do you use Veeam ONE for infrastructure monitoring? There are useful assessment reports helping you to tune your infrastructure.

[SpiderPig]

Hello, Mr.Hendrix

Im guessing the cpu/ram will be shifted when using off-host proxy but the disk-io will remain?
Can i use Veeam One free for this?

Thanks so much

[Shestakov]

You are correct, Off-host proxy will help mostly with Memory performance and CPU, however, when memory pressure is higher than 100% (memory demand is bigger than memory capacity) Hyper-V moves exceeding memory pages to disk.
Yes, the assessment Veeam ONE reports are available in free version.
There is a good blog post about this.

[SpiderPig]

Thanks for guiding me through the "purple haze" of Veeaming, Mr.Hendrix

Have a lovely weekend!

[Shestakov]

You are very welcome!
Once you have a question, don`t hesitate asking.