kernel: pcieport 0000:c0:03.1: PME: Spurious native interrupt!

K

kilping

Member
Proxmox Subscriber
Sep 10, 2020
7
0
6
34
Hi everyone,

I was wondering if there is any fix for this issue. It looks like it's a problem with intel 10g SFP+. I have found a previous post and have used the ethtool work around because like original poster my logs were filling up and bringing down the server. For reference the machine and network adapter I have is a HPE DL385 gen10 with a HP Ethernet 10Gb 2-port 560FLR-SFP+ Adapter, running Proxmox VE 7.1-7



intel_nic.png

https://forum.proxmox.com/threads/pme-spurious-native-interrupt-kernel-meldungen.62850/

nic_issue.png
 
Do you have the latest BIOS installed? Is the firmware of the NIC the latest version?
 
Mira,

Thank you for your response, the BIOS is up to date. The firmware on the NIC is a little more challenging. HPE doesn't make it easy to upgrade firmware unless you are on their supported platforms. I did find a version for Windows Server and found .bin files in the folder labeled 'FW'. This was my work around for a previous NIC I ha, however when I tried this time I could only find a .txt file for my specified NIC. Below is the HPE firmware I tried to use. I may try to boot into fedora from a USB and try installing firmware that way, any thoughts?

https://support.hpe.com/hpesc/public/swd/detail?swItemId=MTX_419d98ca460242a8ada1795c5f#tab3
 
Update:

I was able to get the service pack download from HPE, I have installed all the latest firmware and the issue still persists. As soon as ProxmoxVE booted I went to the syslog and saw the same errors. If i run the command "ethtool -C eno5 rx-usecs 0 && /sbin/ethtool -C eno6 rx-usecs 0" the errors stop. Really want to figure out what is going on. Really putting a damper on my 10Gb update party, lol!
 
As this is the first thread on the issue with a staff reply I thought I'd add the following:

It appears that it's 'fixed' not just by disabling interrupt coalescing but rather toggling it; it can safely be turned back on again after disabling. Also as noticed by 'ajeffco' in another thread on the issue that this toggle only works if done after at least one VM has been booted on the node since Proxmox boot.

It seems to only affect one NIC chipset, Intel 82599ES and has been occurring since at least Proxmox version 6.

Edit: there is an unraid forum thread https://forums.unraid.net/topic/108...tive-interrupt-after-upgrade-from-683-to-692/ about the same issue but not much useful in it except that it seems the linux kernel code logging this error was changed sometime in 2019 to always log to dmesg which would explain this error suddenly showing up.
 
Last edited:
  • Like
Reactions: kilping
csutcliff said:
As this is the first thread on the issue with a staff reply I thought I'd add the following:

It appears that it's 'fixed' not just by disabling interrupt coalescing but rather toggling it; it can safely be turned back on again after disabling. Also as noticed by 'ajeffco' in another thread on the issue that this toggle only works if done after at least one VM has been booted on the node since proxmox boot.

It seems to only affect one NIC chipset, Intel 82599ES and has been occuring since at least Proxmox version 6.
Click to expand...
Thank you for your response here and on Reddit. I've been pulling my hair out trying to find more information. Aslo, please excuse my ignorance but by toggling I assume running the same command as before but just changing the 0 to 1 would put the NICS in the same state correct?

Also worth noting, I have vms that start on boot which is in line with what what the other user was experiencing as well.
 
The usecs parameters specify how many microseconds after at least 1 packet is received/transmitted before generating an interrupt. In my case the default is 3. You can check yours by running ethtool -c <interface> and reading the output on a fresh boot before you disable it.
 
  • Like
Reactions: kilping
It's been a while and this is still an issue. In the mean time I've also discovered that disabling PCIe power management in /etc/kernel/cmdline (pcie_aspm=off) also stops this log spam, with the benefit of not needing manual intervention each boot.
 
  • Like
Reactions: ArcticCode
Thank you for this post. I noticed the same issue today on one of the systems. I work for a company that has 6 systems deployed. I only have one of them showing the issue but it is on the newest software 7.2. The others are on Verizon 5.4. I dont know if that is effecting the issue but it is a little details.

I have checked the hardware and 4 of the 6 have the same card. The systems with the older software is not showing the error. They are due to be upgraded in the next few weeks.

root@####~# lspci -vv -t|grep 03
+-03.2-[07-08]--+-00.0 Intel Corporation 82599ES 10-Gigabit SFI/SFP+ Network Connection
 
Last edited:
Faldiin said:
Thank you for this post. I noticed the same issue today on one of the systems. I work for a company that has 6 systems deployed. I only have one of them showing the issue but it is on the newest software 7.2. The others are on Verizon 5.4. I dont know if that is effecting the issue but it is a little details.

I have checked the hardware and 4 of the 6 have the same card. The systems with the older software is not showing the error. They are due to be upgraded in the next few weeks.

root@####~# lspci -vv -t|grep 03
+-03.2-[07-08]--+-00.0 Intel Corporation 82599ES 10-Gigabit SFI/SFP+ Network Connection
Click to expand...
Did you find a solution to this? I have a Proliant DL380 G9 with an "Intel Corporation 82599ES 10-Gigabit SFI/SFP+ Network Connection" card. I've installed all firmware updates for the board and netcard but no luck.

Until recently running "/usr/sbin/ethtool -C eno50 rx-usecs 0" would work but it no longer seems to. Running kernel 5.15.74-1-pve.

This is driving me crazy as I can't seem to find a solution for it now.

Any help / advice greatly appreciated.
 
May I jump in?
HPE DL 360 Gen 9 (our Lab-Environment) is showing the same behaviour.
lspci -vv | egrep Net
04:00.0 Ethernet controller: Intel Corporation 82599ES 10-Gigabit SFI/SFP+ Network Connection (rev 01)
04:00.1 Ethernet controller: Intel Corporation 82599ES 10-Gigabit SFI/SFP+ Network Connection (rev 01)

ProxMox 8.03
 
Not sure if this will work for others, but I went into the bios settings for the 10gb card and disabled all references to PXE and SRIOV. This seems to have solved the issue and no further errors.

Don't forget to disable shared memory setting also in the network device config - that will break PCIe passthrough on Proxmox if you enable it.
 
You must log in or register to reply here.

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!

Quick Navigation

Home Get Subscription Wiki Downloads Proxmox Customer Portal About

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!
Community platform by XenForo® © 2010-2024 XenForo Ltd.
Top Bottom
Back