stepmback

Great thanks.

I think I figured it out. I had to disable flow control on the nic on my windows PC. That seems to have drastically increase speed. Now getting 950 MB/sec with no drop off. Related. Should I be using Jumbo frames (9014)? What is the downside?

I have been messing with this a while. I have workstation connected with Agg switch SFP+ 10GB to unraid connected SFP+ 10GB. I am copying large files (70GB) from workstation to server. I have this folder going to cache drive (nvme NVME 1TB Samsung pro) and it starts at 200MB a second and other times it does 1.2 GB a second then dies down to 200 mb a second. I was troubleshooting and wondering if it could be the setup of my cache drive? I have plenty of free space (700 GB on nvme). Could it be the "copy on write" set to auto? Any help is appreciated.

I can also run glances but that does not show load (by core) or voltage as best I can tell.

Does not show voltage or load. I can see the load on my main dashboard but not voltage or temp (each core). Hoping there is an app that will show all in one screen.

I can see my core temp on my cpu but is there a plugin that will show me all the temps of each core, voltage, load at once? Something like Open Hardware Monitor.

Thanks. I moved the card and it booted fine. All the drives are showing. Everything is running smooth and VMs started. I even did some testing and am getting 800 MB/s from my workstation to unraid serve (nvme cache is amazing). Getting 180BM from server (array) to workstation but that is expected because of old drives. Have not run a parity check so we will see how that goes in a couple weeks. If I do go with a graphics card I guess I will move the HBA to the x1 slot and live with a couple day parity check. Thanks for your help. My next thing to tinker on is to see if I can actually get my Windows 11 VM to connect at 10gb.

If i move the HBA to the x16 slot would that affect my nvme drive and its speeds etc?

That is the problem. One of the bottom two 16x slots is x1 and the other is x4. I have the HBA in x4 and the SFP in x1. The SFP works but getting like 180MB transfer speeds. I swapped the cards because the HBA was taking days to do a parity check. After doing that I noticed the drop in speeds on the SFP.

I should planned this better. I have an intel i9 14900 chip in a gigabyte motherboard that has 5 PCIE slots. Only one of the slots is x4 and that currently has my HBA card. However, I also have a SFP+ card that needs a x4 slot to get 10gbe. I do have a free x16 slot which I am considering putting my HBA card in however I am concerned that if I do that my NVME will be affected. * I don't plan on gaming in the future but I had wanted to leave the slot open to future proof. Thoughts? Here is a link to the MB I have: https://www.gigabyte.com/Motherboard/Z790-EAGLE-AX-rev-1x#kf mb_manual_z790-eagle-ax_1002_e.pdf

Thanks for finding that. I appreciate all the help. I changed the card to the PCIEX4 slot and it is now showing as 4X

My screwup. I don't have a graphics card so the PCIEX16 is available but that may change which is why I did not use that. I also figured maybe incorrectly that because the card is old there is no point in putting it there. Want to make sure i put it in the right slot. Put it in the PCIEX4?

I stopped the parity check and downloaded diskspeed am running benchmarks now. Anything I should be looking for?

Should I stop the parity check or let it run? Also, in my upgrade build I moved the HBA card to 8x pcie slot (I think it is the 8x, I know not the 16x) slot. Could that be the issue? No enough lanes?