Quantcast
Channel: High Availability (Clustering) forum
Viewing all articles
Browse latest Browse all 3614

SET switch on 1GB adapters maxing core 0

$
0
0

I have a brand new 2016 cluster sitting on 2x DL360 G10 nodes.

2x cpu's of 8 cores each (so 16 cores\32 logical processors)

I've joined all 4x 1GB NICS into the SET Switch.

Created vNIC for LM, Cluster and Management (no SMB, as its direct SAS)

I created a SET switch but whenever i live migrate VM's around, Core 0 maxes out which means i get around 1.6GBit out of the NIC's  (is showing as 4GBit)

I understand VMQ is a big no no for 1GB (i'm used to working on 10Gbit)

I set RSS to move away from core 0, but still does the same.

Set-NetadapterRSS -Name OB1 -BaseProcessorNumber 2 -NumaNode 0
Set-NetadapterRSS -Name OB2 -BaseProcessorNumber 8 -NumaNode 0
Set-NetadapterRSS -Name OB3 -BaseProcessorNumber 2 -NumaNode 1
Set-NetadapterRSS -Name OB4 -BaseProcessorNumber 8 -NumaNode 1

What am i missing?  What have i done wrong?

Thanks 


Viewing all articles
Browse latest Browse all 3614

Trending Articles



<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>