Score:1

Issue with two Dell R620 servers connected directly with Cat5e cable get slow speeds

ve flag

I'm trying to set up a new Hyper-V cluster to run my virtual environment on. Decided to try out 2019 and currently I have two HOST servers (Dell R620s) connected together using a cat5 cable directly (no switch). Each host has a total of 4 NICs (2x10Gb and 2x1Gb). The issue was really bad on one of the 10Gb NICs so I decided to do the tests on the 1Gb NICs on each server.

I have the other NICs connected to other things in my production environment but was getting mad about the slow speeds across the network so I wanted to do some testing...

HOST1 : 192.168.35.1 ...directly connected to... HOST2 : 192.168.35.2

On host1 I try to send over a large VM and this is the results... enter image description here

Any idea what I might have configured wrong to make the network speed so bad? I have also disabled VMQ on all NICs just to test things out. Not sure if this is something with 2019 or maybe the NICs etc.

Thanks for any help!

Andrew Henle avatar
ph flag
What are you NIC settings? Did you try another cable, one that's known to be good? *connected together using a cat5 cable* FWIW, you need cat6 for 10Gbase-T...
ve flag
@AndrewHenle What NIC settings are you after specifically? That's a very vague question. Are you talking about IP settings, hardware settings like VMQ etc? Yes, other cables were tried, and just to troubleshoot I have a strait cable connected from one NIC to the other eliminating the switch as a possible issue. About the Cat6 comment, the strait cable (cat5e) was just for testing and was on a 1Gb, not 10Gb. I just want to get speeds that are realistic for Cat5e though. Something better than 30 MB/s. ;)
Wilson Hauck avatar
jp flag
Could be a book lying on a keyboard with any logged on workstation.
mangohost

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.