PVE 6.0-7 + ixgbe firmware errors

For people who don't think this is causing any issues, I have 2 ports bridged and goes:
192.168.2.100 <=> 192.168.2.1 <=> 192.168.2.200
With B being the machine causing problems.

From 192.168.2.200 a Win10 box before update:
E:\data\iperf3>iperf3.exe -c 192.168.2.1
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-10.00 sec 3.30 GBytes 2.84 Gbits/sec sender
[ 4] 0.00-10.00 sec 3.30 GBytes 2.84 Gbits/sec receiver

E:\data\iperf3>iperf3.exe -c 192.168.2.100
Connecting to host 192.168.2.100, port 5201
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-10.00 sec 280 MBytes 235 Mbits/sec sender
[ 4] 0.00-10.00 sec 280 MBytes 235 Mbits/sec receiver
2.8Gb going to the interface, 235**Mb** going through it, or an order of magnitude less.

After updating driver:
E:\data\iperf3>iperf3.exe -c 192.168.2.100
Connecting to host 192.168.2.100, port 5201
[ ID] Interval Transfer Bandwidth
[ 4] 0.00-10.00 sec 3.37 GBytes 2.90 Gbits/sec sender
[ 4] 0.00-10.00 sec 3.37 GBytes 2.90 Gbits/sec receiver

Going through the bridge is the same speed as going to the bridge node. Thanks so much for these instructions!
 
  • Like
Reactions: stark
Just a note that this does not build with Proxmox 6.2 and the errors are back. The most recent driver from Intel is only tested up to kernel 5.3 not 5.4. There's a few compile errors that I'm trying to work through.
 
Last edited:
Just a note that this does not build with Proxmox 6.1 and the errors are back. The most recent driver from Intel is only tested up to kernel 5.3 not 5.4. There's a few compile errors that I'm trying to work through.

You mean that it doesn't build with Proxmox 6.2? Yeah, that's unfortunate but we'll have to wait on Intel to update their driver to work with kernel 5.4.

I'm hoping that it won't be too long. Until then I'm going to force myself to put up with the errors.
 
You mean that it doesn't build with Proxmox 6.2? Yeah, that's unfortunate but we'll have to wait on Intel to update their driver to work with kernel 5.4.

I'm hoping that it won't be too long. Until then I'm going to force myself to put up with the errors.

Yes, I updated my post.
 
Ok, I have a patch that's been running for about an hour or so....
 

Attachments

  • linux-5-4.txt
    4.5 KB · Views: 31
Let me know how it works, between that and WSL2 the iperf3 report up thread went from 3g to 10g
That's a pretty amazing increase. I did a quick iperf2 test before and after and the results were similar ~ 9.86Gb/sec for each one. No massive increase for me.. :( :)
 
Last edited:
Sorry to bump an old thread, this issue looks to be back on the latest kernel + driver:

Bash:
root@proxmox1:~# pveversion
pve-manager/8.1.4/ec5affc9e41f1d79 (running kernel: 6.5.11-7-pve)

Bash:
root@proxmox1:~# ethtool -i ens1f0
driver: ixgbe
version: 5.19.9
firmware-version: 0x800001a5
expansion-rom-version:
bus-info: 0000:07:00.0
supports-statistics: yes
supports-test: yes
supports-eeprom-access: yes
supports-register-dump: yes
supports-priv-flags: yes

dmesg is showing:

Bash:
...
[ 1523.104064] ixgbe 0000:07:00.0: Warning firmware error detected FWSM: 0x00000000
[ 1525.120072] ixgbe 0000:07:00.0: Warning firmware error detected FWSM: 0x00000000
[ 1527.136088] ixgbe 0000:07:00.0: Warning firmware error detected FWSM: 0x00000000
...

This is with a Dell branded X520-T2
Anyone else seeing this?
 
Last edited:
Had that happen sometimes and it be cleared by a reseat / wipe down of the PCIe pins with isopropyl alcohol.

Give that a try and see if it clears up.
 
Had that happen sometimes and it be cleared by a reseat / wipe down of the PCIe pins with isopropyl alcohol.

Give that a try and see if it clears up.
It looks like TrueNAS team found this is likely a bug in the Intel firmware not setting a bit after boot: https://ixsystems.atlassian.net/browse/NAS-121174

There's a edit they did to the Intel driver to suppress the messages after the first one here: https://github.com/truenas/linux/pull/81/commits/b42356c9b0419af14df988c9970e667e91cc075a
 
Had a quick read of that earlier today, sounds like they silenced the repeated warnings rather than fixing a bug per se. Seems unclear if this is a red herring error warning or indicative of any real issues (probably worth you doing some iperf3 tests to see if you have any poor performance.)

Edit: probably also worth double checking the error warnings are only appearing in the latest kernel too and looking through the git history for the driver to look for anything obvious.

It would seem a sensible patch to include in Proxmox, but won't address the root cause of the firmware error (whether this is a red herring or not.)
 
Last edited:

About

The Proxmox community has been around for many years and offers help and support for Proxmox VE, Proxmox Backup Server, and Proxmox Mail Gateway.
We think our community is one of the best thanks to people like you!

Get your subscription!

The Proxmox team works very hard to make sure you are running the best software and getting stable updates and security enhancements, as well as quick enterprise support. Tens of thousands of happy customers have a Proxmox subscription. Get yours easily in our online shop.

Buy now!