"PME: Spurious native interrupt!" Kernel Meldungen

smasty

Member
Mar 21, 2019
9
7
8
Hallo zusammen,

nachdem ich nun den ersten PVE-Knoten von 6.0.9 auf 6.1.5 (und damit Kernel 5.3.13) angehoben habe, wird mein Syslog mit folgenden Meldungen kontinuierlich befüllt:

[...]
Jan 6 10:00:01 kernel: [ 8076.840412] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:04 kernel: [ 8078.958690] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:06 kernel: [ 8080.971962] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:06 kernel: [ 8080.988809] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:09 kernel: [ 8084.103786] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:09 kernel: [ 8084.103963] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:11 kernel: [ 8086.457771] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:16 kernel: [ 8091.345566] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:16 kernel: [ 8091.353984] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:26 kernel: [ 8101.770559] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:27 kernel: [ 8102.724566] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:27 kernel: [ 8102.869279] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:28 kernel: [ 8103.431608] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:31 kernel: [ 8106.841754] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:31 kernel: [ 8106.856727] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:33 kernel: [ 8108.399521] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:38 kernel: [ 8113.399608] pcieport 0000:00:03.0: PME: Spurious native interrupt!
Jan 6 10:00:39 kernel: [ 8114.350334] pcieport 0000:00:03.0: PME: Spurious native interrupt!
[...]

Wie man an den zeitlichen Abständen sieht, kommen da schon viele Log Messages zusammen...

Es geht dabei wohl um folgendes PCI-Device:

# lspci -vv -t|grep 03.0
| +-03.0-[83]--
+-03.0-[04-05]--+-00.0 Intel Corporation 82599ES 10-Gigabit SFI/SFP+ Network Connection

Die Meldungen erscheinen allerdings erst im Log, nachdem ich wieder eine VM zurückmigriert habe bzw. dort eine VM läuft.

Ich vermute, dass sich das mit dem/den nächsten Kernel Updates erledigen wird, hat hier jemand ähnliches im Log?
 

rdfeij

Member
Jan 8, 2020
6
1
8
43
Same problem here. Since update yesterday from version 5.4.x to 6.1-5
We have the same Intel 10Gbit interfaces.
We had over 64GB of logging in kern.log filling up OS disk
 

rdfeij

Member
Jan 8, 2020
6
1
8
43
Same problem here. Since update yesterday from version 5.4.x to 6.1-5
We have the same Intel 10Gbit interfaces.
We had over 64GB of logging in kern.log filling up OS disk

This morning our 3th node crashed again (out of diskspace)
After deleting:
/var/log/syslog
/var/log/messages
/var/log/kern.log
and a reboot it is working again.

The spurious interrupt records are filling up logging very fast.
For now solved by creating a cronjob emptying these files every hour.
 

spyfly

Active Member
Jul 22, 2017
30
0
31
22
I am experiencing the same issue with the same network card, as seen below.
Code:
lspci -vv -t|grep 03.2
             +-03.2-[07]--+-00.0  Intel Corporation 82599ES 10-Gigabit SFI/SFP+ Network Connection

My syslog looks like this as well:
Code:
[...]
Apr 14 20:43:02 *** kernel: [2413144.982743] pcieport 0000:00:03.2: PME: Spurious native interrupt!
Apr 14 20:43:07 *** kernel: [2413150.297445] pcieport 0000:00:03.2: PME: Spurious native interrupt!
Apr 14 20:43:07 *** kernel: [2413150.505500] pcieport 0000:00:03.2: PME: Spurious native interrupt!
Apr 14 20:43:18 *** kernel: [2413161.499213] pcieport 0000:00:03.2: PME: Spurious native interrupt!
[...]
 
Jul 10, 2020
12
6
8
33
Hi,

I have the same issue atm with the 82599ES driver on proxmox. I've found that if you disable interrupt coalescing on the affected cards then the errors stop happening.

sudo ethtool -C <iface> rx-usecs 0
 

NTS

Member
Apr 21, 2020
11
0
6
27
Hi
We have the same issue:
Bash:
root@proxmox1:/etc/rsyslog.d# lspci
...
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)
...

dmesg:
Bash:
[Thu Jul 16 15:45:32 2020] pcieport 0000:00:02.2: PME: Spurious native interrupt!

We run Virtual Environment 6.2-10 since today, and still got this messages
any ideas, or does we have run the ethtool command like above?

@csutcliff - did you got any issues after this command ?
 
Jul 10, 2020
12
6
8
33
Hi
We have the same issue:
Bash:
root@proxmox1:/etc/rsyslog.d# lspci
...
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)
...

dmesg:
Bash:
[Thu Jul 16 15:45:32 2020] pcieport 0000:00:02.2: PME: Spurious native interrupt!

We run Virtual Environment 6.2-10 since today, and still got this messages
any ideas, or does we have run the ethtool command like above?

@csutcliff - did you got any issues after this command ?

I haven't had any issue since turning off interrupt coalescing. The expected side effect of doing this is slight cpu usage increase as there will be more interrupts. In my environment this is immeasurably small.
 

rdfeij

Member
Jan 8, 2020
6
1
8
43
Hi
We have the same issue:
Bash:
root@proxmox1:/etc/rsyslog.d# lspci
...
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)
...

dmesg:
Bash:
[Thu Jul 16 15:45:32 2020] pcieport 0000:00:02.2: PME: Spurious native interrupt!

We run Virtual Environment 6.2-10 since today, and still got this messages
any ideas, or does we have run the ethtool command like above?

@csutcliff - did you got any issues after this command ?

Run the command above. Keep in mind the interface will reinitialize and disconnect for a few seconds.
It worked for our enviroment. CPU raise not noticable.
 

NTS

Member
Apr 21, 2020
11
0
6
27
Run the command above. Keep in mind the interface will reinitialize and disconnect for a few seconds.
It worked for our enviroment. CPU raise not noticable.
Hi Rdfeji
Did you ran the command below on the physical or bond interface?
Bash:
sudo ethtool -C <iface> rx-usecs 0
 

NTS

Member
Apr 21, 2020
11
0
6
27
fine, @reboot is straight forward.
i did a script as follows - but untested yet!
Bash:
root@proxmox1:/etc/network/if-pre-up.d# cat ethtool-rx-usecs
#!/bin/sh
ETHTOOL=/usr/sbin/ethtool
$ETHTOOL -C eno49 rx-usecs 0
$ETHTOOL -C eno50 rx-usecs 0
And of course, change interfaces as needed
 

rdfeij

Member
Jan 8, 2020
6
1
8
43
Funny, i'm still investigating. Last weekend i updated all nodes.
I still have the mentioned temporary fix active (run with Cron @reboot)and after a reboot i saw "some but not many" spurious logs again.
Trying to manual disable interrupts it said "nothing to change" so Cron is working.
After enabling interupts again spurious messages were gone?????
Having to investigate this further and still awaiting fix????
 
  • Like
Reactions: kilping

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 your own in 60 seconds.

Buy now!