Best way to change faulty pcie ethernet card without breaking configuration

My network card is dying and i don’t know how to properly change it. Because if I just shutdown and change it hw-id of new card will be different ant it will go up as eth2 instead of old eth1.
Thinking about adding new card and just rewriting all configuration for new interface and then load that configuration.

dmesg -T is below

[Fri Jan 28 18:57:31 2022] 8139too 0000:03:00.0 eth1: link down
[Fri Jan 28 18:57:33 2022] 8139too 0000:03:00.0 eth1: link up, 100Mbps, full-duplex, lpa 0x45E1
[Fri Jan 28 18:57:55 2022] 8139too 0000:03:00.0 eth1: link down
[Fri Jan 28 18:57:57 2022] 8139too 0000:03:00.0 eth1: link up, 100Mbps, full-duplex, lpa 0x45E1
[Fri Jan 28 18:58:05 2022] 8139too 0000:03:00.0 eth1: link down
[Fri Jan 28 18:58:07 2022] 8139too 0000:03:00.0 eth1: link up, 100Mbps, full-duplex, lpa 0x41E1
[Fri Jan 28 19:11:31 2022] 8139too 0000:03:00.0 eth1: link down
[Fri Jan 28 19:11:33 2022] 8139too 0000:03:00.0 eth1: link up, 100Mbps, full-duplex, lpa 0x45E1
[Fri Jan 28 19:11:55 2022] 8139too 0000:03:00.0 eth1: link down
[Fri Jan 28 19:11:57 2022] 8139too 0000:03:00.0 eth1: link up, 100Mbps, full-duplex, lpa 0x45E1
[Fri Jan 28 19:12:01 2022] 8139too 0000:03:00.0 eth1: link down
[Fri Jan 28 19:12:03 2022] 8139too 0000:03:00.0 eth1: link up, 100Mbps, full-duplex, lpa 0x41E1
[Fri Jan 28 19:27:34 2022] 8139too 0000:03:00.0 eth1: link down
[Fri Jan 28 19:27:36 2022] 8139too 0000:03:00.0 eth1: link up, 100Mbps, full-duplex, lpa 0x45E1
[Fri Jan 28 19:27:59 2022] 8139too 0000:03:00.0 eth1: link down
[Fri Jan 28 19:28:00 2022] 8139too 0000:03:00.0 eth1: link up, 100Mbps, full-duplex, lpa 0x45E1
[Fri Jan 28 19:28:07 2022] 8139too 0000:03:00.0 eth1: link down
[Fri Jan 28 19:28:09 2022] 8139too 0000:03:00.0 eth1: link up, 100Mbps, full-duplex, lpa 0x41E1
[Fri Jan 28 20:06:50 2022] 8139too 0000:03:00.0 eth1: link down
[Fri Jan 28 20:06:52 2022] 8139too 0000:03:00.0 eth1: link up, 100Mbps, full-duplex, lpa 0x45E1
[Fri Jan 28 20:07:14 2022] 8139too 0000:03:00.0 eth1: link down
[Fri Jan 28 20:07:15 2022] 8139too 0000:03:00.0 eth1: link up, 100Mbps, full-duplex, lpa 0x45E1
[Fri Jan 28 20:07:22 2022] 8139too 0000:03:00.0 eth1: link down
[Fri Jan 28 20:07:24 2022] 8139too 0000:03:00.0 eth1: link up, 100Mbps, full-duplex, lpa 0x41E1
[Sat Jan 29 00:00:04 2022] 8139too 0000:03:00.0 eth1: link down
[Sat Jan 29 00:00:06 2022] 8139too 0000:03:00.0 eth1: link up, 100Mbps, full-duplex, lpa 0x45E1
[Sat Jan 29 00:00:28 2022] 8139too 0000:03:00.0 eth1: link down

At the time when this happens I see huge spike in latency to host(like 900ms or more)

Note down the mac address of the new card, apply it as the hw-id of the interface you’re replacing, shutdown the machine, replace card, and when you boot up it should all be OK in theory.

2 Likes

This topic was automatically closed 2 days after the last reply. New replies are no longer allowed.