1 person found this helpful
Please take a look at https://kb.vmware.com/s/article/1004048 (scoll down to the HP specific section) to see whether this is what you are looking for.
PS: Only static trunks are supported with Standard vSwitches.
I will look into these links and post back. Thank you.
I took a look at the article André provided and it helped me quite a bit. Thank you for that. There is one problem I think I am going to have some trouble overcoming. In the article, the HP switch ports are expected to be set to trunk mode. This model of switch doesn't appear to have this option in the configuration. The options I am presented with are only lacp dynamic and static. I set the two ports to static and this seemed to satisfy ESXI. It seems the ability to set in TRUNK is not available in the JH295A. I didn't see it in the CLI either. Doesn't mean it's not there.
I then created a virtual switch with 2 x uplinks (both of the ports on the physical nic) with the following settings:
Then I created the following port group:
After these were created, I removed the existing network adapter from my Windows Server VM the created a new VMXNET3 network adapter in the VM settings and installed VMWare tools. My VM Server shows that I have a 10Gb uplink status but based on some testing results, It doesn't appear that I am operating under trunked 10Gb connections.
Granted, our server backbone may be 10Gb but our main network switching is 1Gb. Our test consisted of 3 users simultaneously copying a large file from the VM to their computers. During a single copy, we reached file transfer speeds of 112MB/sec, when we performed the simultaneous copies, the file transfer speed just split up into thirds. My understanding of trunking is that this shouldn't be the case. Since there are two routes in/out, the connections are supposed to split allowing more bandwith and I would expect that we should see better performance. Is my understanding of trunking wrong? Shouldn't at least two copies be able to happen at full speed?
I know that perfmon is a better test for speeds and I have used it in other situations, but I don't think it is what I need in this instance.
I got this working as expected. See post with images of settings. It was working as it should have been, I finally realized that I only have 1Gb uplinks between our main switches so I should not expect more than that through them.
Thanks for the links to the documentation. I will also add that getting a semi-decent understanding of the ESXI terminology was helpful: https://www.vmware.com/content/dam/digitalmarketing/vmware/en/pdf/techpaper/virtual_networking_concepts.pdf
SIDENOTE: HP has now apparently adopted the term trunking to what they used to refer to as VLAN Tagging. Here is a snippet from their Latest HP 1950 manual: Trunk - A trunk port can forward packets from multiple VLANs. Except packets from the port VLAN ID (PVID), packets sent out of a trunk port are VLAN-tagged. Assign a trunk port to the untagged port list of the PVID of the port, and to the tagged port lists of other VLANs
Thanks for the feedback. In case you didn't do that before, ESXi contains a handy command line utility esxtop, which you can use to monitor different things including live throughput on vmnics.