i40e module for X710 Intel NIC

Ataraxxia

New Member
Jul 10, 2023
4
0
1
Hello everyone,

I have come across very peculiar issue. Some time ago my institution acquired 6 identical Dell r7515 servers. I've provisioned them into two Proxmox clusters 3 nodes each. They all have 10Gbit Intel X710 NIC, dual port, aggregated into LACP bond.
1st cluster works no problem, the second one however has upload capped at around 6Mbit/s.
All servers are running the same proxmox version and kernel. All have the same firmware, bios, etc.

The working cluster initializes seemingly without any issues:

Code:
root@ocean-pve1-r7515:~# dmesg -T | grep i40e
[Tue Dec 17 16:54:44 2024] i40e: Intel(R) Ethernet Connection XL710 Network Driver
[Tue Dec 17 16:54:44 2024] i40e: Copyright (c) 2013 - 2019 Intel Corporation.
[Tue Dec 17 16:54:44 2024] i40e 0000:02:00.0: fw 9.840.76614 api 1.15 nvm 9.40 0x8000e9c2 22.5.7 [8086:1572] [8086:0006]
[Tue Dec 17 16:54:44 2024] i40e 0000:02:00.0: MAC address: 6c:fe:54:7b:e9:20
[Tue Dec 17 16:54:44 2024] i40e 0000:02:00.0: FW LLDP is enabled
[Tue Dec 17 16:54:44 2024] i40e 0000:02:00.0 eth2: NIC Link is Up, 10 Gbps Full Duplex, Flow Control: None
[Tue Dec 17 16:54:44 2024] i40e 0000:02:00.0: PCI-Express: Speed 8.0GT/s Width x8
[Tue Dec 17 16:54:44 2024] i40e 0000:02:00.0: Features: PF-id[0] VFs: 64 VSIs: 66 QP: 48 RSS FD_ATR FD_SB NTUPLE DCB VxLAN Geneve PTP VEPA
[Tue Dec 17 16:54:44 2024] i40e 0000:02:00.1: fw 9.840.76614 api 1.15 nvm 9.40 0x8000e9c2 22.5.7 [8086:1572] [8086:0006]
[Tue Dec 17 16:54:44 2024] i40e 0000:02:00.1: MAC address: 6c:fe:54:7b:e9:21
[Tue Dec 17 16:54:44 2024] i40e 0000:02:00.1: FW LLDP is enabled
[Tue Dec 17 16:54:44 2024] i40e 0000:02:00.1 eth3: NIC Link is Up, 10 Gbps Full Duplex, Flow Control: None
[Tue Dec 17 16:54:44 2024] i40e 0000:02:00.1: PCI-Express: Speed 8.0GT/s Width x8
[Tue Dec 17 16:54:44 2024] i40e 0000:02:00.1: Features: PF-id[1] VFs: 64 VSIs: 66 QP: 48 RSS FD_ATR FD_SB NTUPLE DCB VxLAN Geneve PTP VEPA
[Tue Dec 17 16:54:45 2024] i40e 0000:02:00.1 ens5f1np1: renamed from eth3
[Tue Dec 17 16:54:45 2024] i40e 0000:02:00.0 ens5f0np0: renamed from eth2
[Tue Dec 17 16:54:58 2024] i40e 0000:02:00.1 ens5f1np1: set new mac address 6c:fe:54:7b:e9:20
[Tue Dec 17 16:54:58 2024] i40e 0000:02:00.0 ens5f0np0: entered allmulticast mode
[Tue Dec 17 16:54:58 2024] i40e 0000:02:00.1 ens5f1np1: entered allmulticast mode
[Tue Dec 17 16:54:58 2024] i40e 0000:02:00.0 ens5f0np0: entered promiscuous mode
[Tue Dec 17 16:54:58 2024] i40e 0000:02:00.1 ens5f1np1: entered promiscuous mode
[Tue Dec 17 16:54:58 2024] i40e 0000:02:00.0: entering allmulti mode.
[Tue Dec 17 16:54:58 2024] i40e 0000:02:00.1: entering allmulti mode.

But on the other cluster the dmesg has issues (this might seem it has newer FW because i tried recent Dell FW upgrade from 22.x to 23.x for the NIC today, but the result is all the same):

Code:
root@paw1-pve2-r7515:~# dmesg -T | grep i40e
[Fri Jan 17 17:58:10 2025] i40e: Intel(R) Ethernet Connection XL710 Network Driver
[Fri Jan 17 17:58:10 2025] i40e: Copyright (c) 2013 - 2019 Intel Corporation.
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.0: fw 9.851.77832 api 1.15 nvm 9.50 0x8000f255 23.0.8 [8086:1572] [8086:0006]
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.0: MAC address: 6c:fe:54:7b:fe:00
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.0: FW LLDP is enabled
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.0: couldn't add VEB, err -EIO aq_err I40E_AQ_RC_EINVAL
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.0: couldn't add VEB
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.0: Couldn't create FDir VSI
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.0 eth2: NIC Link is Up, 10 Gbps Full Duplex, Flow Control: None
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.0: PCI-Express: Speed 8.0GT/s Width x8
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.0: Features: PF-id[0] VFs: 64 VSIs: 2 QP: 48 RSS FD_ATR DCB VxLAN Geneve PTP VEPA
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.1: fw 9.851.77832 api 1.15 nvm 9.50 0x8000f255 23.0.8 [8086:1572] [8086:0006]
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.1: MAC address: 6c:fe:54:7b:fe:01
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.1: FW LLDP is enabled
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.1: couldn't add VEB, err -EIO aq_err I40E_AQ_RC_EINVAL
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.1: couldn't add VEB
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.1: Couldn't create FDir VSI
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.1 eth3: NIC Link is Up, 10 Gbps Full Duplex, Flow Control: None
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.1: PCI-Express: Speed 8.0GT/s Width x8
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.1: Features: PF-id[1] VFs: 64 VSIs: 2 QP: 48 RSS FD_ATR DCB VxLAN Geneve PTP VEPA
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.0 ens5f0np0: renamed from eth2
[Fri Jan 17 17:58:10 2025] i40e 0000:02:00.1 ens5f1np1: renamed from eth3
[Fri Jan 17 17:58:23 2025] i40e 0000:02:00.1 ens5f1np1: set new mac address 6c:fe:54:7b:fe:00
[Fri Jan 17 17:58:23 2025] i40e 0000:02:00.0 ens5f0np0: entered allmulticast mode
[Fri Jan 17 17:58:23 2025] i40e 0000:02:00.1 ens5f1np1: entered allmulticast mode
[Fri Jan 17 17:58:23 2025] i40e 0000:02:00.0 ens5f0np0: entered promiscuous mode
[Fri Jan 17 17:58:23 2025] i40e 0000:02:00.1 ens5f1np1: entered promiscuous mode
[Fri Jan 17 17:58:23 2025] i40e 0000:02:00.1: entering allmulti mode.
[Fri Jan 17 17:58:24 2025] i40e 0000:02:00.0: entering allmulti mode.
[Fri Jan 17 17:58:28 2025] i40e 0000:02:00.1 ens5f1np1: NETDEV WATCHDOG: CPU: 13: transmit queue 23 timed out 5160 ms
[Fri Jan 17 17:58:28 2025] i40e 0000:02:00.1 ens5f1np1: tx_timeout: VSI_seid: 389, Q 23, NTC: 0x0, HWB: 0x0, NTU: 0x1, TAIL: 0x1, INT: 0x1
[Fri Jan 17 17:58:28 2025] i40e 0000:02:00.1 ens5f1np1: tx_timeout recovery level 1, txqueue 23
[Fri Jan 17 17:58:28 2025] i40e 0000:02:00.0 ens5f0np0: NETDEV WATCHDOG: CPU: 13: transmit queue 15 timed out 5232 ms
[Fri Jan 17 17:58:28 2025] i40e 0000:02:00.0 ens5f0np0: tx_timeout: VSI_seid: 388, Q 15, NTC: 0x0, HWB: 0x0, NTU: 0x1, TAIL: 0x1, INT: 0x1
[Fri Jan 17 17:58:28 2025] i40e 0000:02:00.0 ens5f0np0: tx_timeout recovery level 1, txqueue 15
[Fri Jan 17 17:58:28 2025] i40e 0000:02:00.1: VSI seid 389 Tx ring 0 disable timeout
[Fri Jan 17 17:58:28 2025] i40e 0000:02:00.0: VSI seid 388 Tx ring 0 disable timeout
[Fri Jan 17 17:58:28 2025] i40e 0000:02:00.1: entering allmulti mode.
[Fri Jan 17 17:58:28 2025] i40e 0000:02:00.0: Error I40E_AQ_RC_EINVAL adding RX filters on PF, promiscuous mode forced on
[Fri Jan 17 17:58:28 2025] i40e 0000:02:00.0: entering allmulti mode.
[Fri Jan 17 17:58:34 2025] i40e 0000:02:00.0 ens5f0np0: NETDEV WATCHDOG: CPU: 13: transmit queue 23 timed out 5688 ms
[Fri Jan 17 17:58:34 2025] i40e 0000:02:00.0 ens5f0np0: tx_timeout: VSI_seid: 388, Q 23, NTC: 0x0, HWB: 0x0, NTU: 0x1, TAIL: 0x1, INT: 0x1
[Fri Jan 17 17:58:34 2025] i40e 0000:02:00.0 ens5f0np0: tx_timeout recovery level 2, txqueue 23
[Fri Jan 17 17:58:34 2025] i40e 0000:02:00.1 ens5f1np1: NETDEV WATCHDOG: CPU: 13: transmit queue 3 timed out 5504 ms
[Fri Jan 17 17:58:34 2025] i40e 0000:02:00.1 ens5f1np1: tx_timeout: VSI_seid: 389, Q 3, NTC: 0x0, HWB: 0x0, NTU: 0x1, TAIL: 0x1, INT: 0x1
[Fri Jan 17 17:58:34 2025] i40e 0000:02:00.1 ens5f1np1: tx_timeout recovery level 2, txqueue 3
[Fri Jan 17 17:58:34 2025] i40e 0000:02:00.0: VSI seid 388 Tx ring 0 disable timeout
[Fri Jan 17 17:58:34 2025] i40e 0000:02:00.1: VSI seid 389 Tx ring 0 disable timeout
[Fri Jan 17 17:58:37 2025] i40e 0000:02:00.0: Error I40E_AQ_RC_EINVAL adding RX filters on PF, promiscuous mode forced on
[Fri Jan 17 17:58:37 2025] i40e 0000:02:00.0: entering allmulti mode.
[Fri Jan 17 17:58:37 2025] i40e 0000:02:00.1: entering allmulti mode.
[Fri Jan 17 17:58:41 2025] i40e 0000:02:00.0: Error I40E_AQ_RC_EINVAL adding RX filters on PF, promiscuous mode forced on
[Fri Jan 17 17:58:41 2025] i40e 0000:02:00.0: entering allmulti mode.
[Fri Jan 17 17:58:41 2025] i40e 0000:02:00.1: entering allmulti mode.

I've tried running other OS'es (Rockylinux 8, 9; Debian 12) and even without network config they have similar dmesg messages.
I have also tried Intel's i40e driver but it only made things worse since with them I lost all ability to connect to the network, no pinging gateway, no nothing.
 

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!