Interfaces (p0/p1/pf0hpf/pf1hpf) missing after reinstall BFB image and reboot on the Bluefield2

Infrastructure & Networking BlueField
1

Hi,

I previously made incorrect configurations on the OVS bridges, resulting in the interfaces p0/p1 and pf0hpf/pf1hpf missing. I have attempted various solutions, including changing the Bluefield2 mode from separated mode to embedded mode and vice versa, rebooting the Bluefield2, and reinstalling the BFB image on the Bluefield2. Despite trying these approaches, the interfaces continue to remain missing.

Currently, the output of “ip link” in Bluefield2 is as follows, which does not contain interfaces such as p0/p1 and pf0hpf/pf1hpf. I wonder why they are still missing even after the reinstallation of the BFB image and reboot, and how I can solve this issue.

ubuntu@localhost:~$ ip link
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: tmfifo_net0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP mode DEFAULT group default qlen 1000
link/ether 00:1a:ca:ff:ff:01 brd ff:ff:ff:ff:ff:ff
3: oob_net0: <NO-CARRIER,BROADCAST,MULTICAST,UP> mtu 1500 qdisc fq_codel state DOWN mode DEFAULT group default qlen 1000
link/ether b8:ce:f6:d1:3e:16 brd ff:ff:ff:ff:ff:ff
4: ibp3s0f0: <BROADCAST,MULTICAST> mtu 4092 qdisc noop state DOWN mode DEFAULT group default qlen 256
link/infiniband 00:00:11:c8:fe:80:00:00:00:00:00:00:b8:ce:f6:03:00:d1:3e:10 brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff
5: ibp3s0f1: <BROADCAST,MULTICAST> mtu 4092 qdisc noop state DOWN mode DEFAULT group default qlen 256
link/infiniband 00:00:10:88:fe:80:00:00:00:00:00:00:b8:ce:f6:03:00:d1:3e:11 brd 00:ff:ff:ff:ff:12:40:1b:ff:ff:00:00:00:00:00:00:ff:ff:ff:ff

Currently, the output of “ovs-vsctl show” in Bluefield2 is as below. It seems like the reinstallation of BFB image has cleared the old ones, but OVS bridges were not created by default upon the first boot after the installation.

ubuntu@localhost:~$ ovs-vsctl show
2e1063cf-1aab-4425-b4fe-9959ab1d2d72
ovs_version: “2.15.1”

And I am concerned that after I create OVS bridges manually, the output of “ovs-vsctl show” will contain errors stating that the network devices p0/p1/pf0hpf/pf1hpf cannot be opened, like below:

Port pf0hpf
Interface pf0hpf
error: “could not open network device pf0hpf (No such device)”

This issue I encountered is pretty like this post: OVS cannot detect the bluefield interface, but the bluefield card works correctly. Thanks in advance for any assistance or suggestions.

2

Hi @hxp0308,

Thanks for your question.
Can you make sure both interfaces are in ethernet mode?
i.e.
root@l-csi-bf2-16:/home/ubuntu# mlxconfig -d 03:00.0 -e q | grep LINK_TYPE

  •    LINK_TYPE_P1                                IB(1)           ETH(2)          ETH(2)          

  •    LINK_TYPE_P2                                IB(1)           ETH(2)          ETH(2)

root@l-csi-bf2-16:/home/ubuntu#

Thanks,
Anatoly

3

Hi Anatoly,

Thank you very much for your help. I ran the command and the output is as below:

root@localhost:/home/ubuntu# mlxconfig -d 03:00.0 -e q | grep LINK_TYPE

  •    LINK_TYPE_P1                        IB(1)           IB(1)           ETH(2)          

  •    LINK_TYPE_P2                        IB(1)           IB(1)           ETH(2)

I wonder if this means the mode of my BF2 interfaces is infiniband, and if I should change it to ethernet mode to make the interfaces appear in the output of ‘ip link’ or ‘ifconfig’. Thank you very much for your time and help!

4

Hi,

The output you are getting means that you’ve configured both links as ethernet, but you need to perform a cold boot of the server to apply the changes:

The meaning of each column:
Configurations: Default Current Next Boot

  •    LINK_TYPE_P1                                IB(1)           ETH(2)          ETH(2)          

  •    LINK_TYPE_P2                                IB(1)           ETH(2)          ETH(2)

root@l-csi-bf2-16:/home/ubuntu#

5

Hi,

Thank you very much for your time! The problem is solved after a cold boot.

6

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.