You are not logged in.
Pages: 1
Apologies in advance for any typos - I'm having to type this on my phone.
One of the NICs on my server/router is no longer working.
Sometimes it's as if it simply isn't there and other times I get the following error:
***INVALID CHECKSUM 00df ***
*** EEPROM MAC address is invalid.
3c59x: vortex_probe1 fails. Returns -22
3c59x: probe of 0000.03.09.0 failed with error -22
By "sometimes" I mean on different boot attempts.
My fear is that this is a hardware error and, since the relevant NIC is an on-board one, I'm concerned that the mobo is on the way out...
Offline
Did you try acpi=off? The internet is filled with a lot of people reporting the same issue
Offline
Ah, yes. I should've mentioned that. I did try acpi=off and it didn't help.
I googled quite a lot for this, but most threads seemed from years ago and not particularly relevant.
Offline
Am I being melodramatic in assuming that, just because the on-board NIC is dying/dead that this means the mobo is dying? Do I just need to put a new NIC in the machine?
Offline
What does ifconfig report? There are tools to change the MAC address. But if the EEPROM is failing could be several things.
I may have to CONSOLE you about your usage of ridiculously easy graphical interfaces...
Look ma, no mouse.
Offline
ifconfig only reports two interfaces: eth1 and lo.
The missing NIC is configured to be eth0 based on the wiki instructions for consistent naming of ethernet devices. If I could get this working without a new NIC I'd be really pleased. As I said above, this is depriving us of internet in our house due to the box being a router.
Offline
You could try ip link to see if there is even a MAC address reported, and possibly use ethtool to reset the interface.
I may have to CONSOLE you about your usage of ridiculously easy graphical interfaces...
Look ma, no mouse.
Offline
Most boots, nothing is reported in ifconfig, ip link, dmesg or lspci. On these occasions there is no error message, or indeed anything at all.
Next time I get a boot with the error, I'll try ip link to see what is reported.
Offline
Tried ip link and others, but I couldn't get any MAC information. In most cases the NIC simply wasn't listed. It was only in dmesg and lspci, but there wasn't any MAC address listed there.
Offline
If you can't do anything with extra/macchanger then I'd say it's hosed.
I may have to CONSOLE you about your usage of ridiculously easy graphical interfaces...
Look ma, no mouse.
Offline
If the on-board is dead, am I right in thinking that doesn't necessarily mean the mobo is dying? In which case I could just put a new NIC in the box.
Offline
Just opened up the box and spotted that it isn't the on-board that's the problem. So, as long as it's not the socket, replacing the card should fix it!
Here's hoping!
Offline
Ok, in desperation I tried just removing and then putting back the NIC in question. All NICs now appear to be losing without error and the MAC address of the problematic card is: 00:a0:24:5b:34:b0
Now, however, netctl profiles will not start. journalctl -xe suggests that a dependency has failed - namely:
Job sys-subsystem-net-devices-eth0.device/start failed with result 'timeout'.
A similar error occurs with both interfaces.
Offline
I fixed that problem based on this post. I now have a minor name resolution issue - but that's another problem and I can probably figure that out.
I'm just so baffled that a card stopped working and then started working just by removing it and putting it back in...
To be honest, I was surprised that the card would fail in the first place. After all, there are no moving parts!
Offline
Pages: 1