rename2 & rename3 devices?

Hi,

I have rename2 & rename3 NICs appearing, and they don’t work, and I can’t get at the devices they are “renames” of. Any advice for how to fix?

Alternatively (if this is simpler), if I want to wipe all the network definitions and start again, what files should I delete & what lines / sections from other files should I clear?

I’m on OS 12.2 64bit. It’s a headless server & has 2 x NICs (1Gb & 100Mb).

History: intermittently bad NIC, so I set up bonding between the onboard NIC & the 1Gb card. Seemed to work OK. A week or so later, replaced bad NIC with one of same brand. Since then, bonding hasn’t worked properly, & these “renamed” devices have been appearing.

Thanks
David

On 04/07/2013 07:36 AM, jetojedno wrote:
>
> Hi,
>
> I have rename2 & rename3 NICs appearing, and they don’t work, and I
> can’t get at the devices they are “renames” of. Any advice for how to
> fix?
>
> Alternatively (if this is simpler), if I want to wipe all the network
> definitions and start again, what files should I delete & what lines /
> sections from other files should I clear?
>
> I’m on OS 12.2 64bit. It’s a headless server & has 2 x NICs (1Gb &
> 100Mb).
>
> History: intermittently bad NIC, so I set up bonding between the
> onboard NIC & the 1Gb card. Seemed to work OK. A week or so later,
> replaced bad NIC with one of same brand. Since then, bonding hasn’t
> worked properly, & these “renamed” devices have been appearing.

The latest versions of udev do not handle renaming of a device to itself the way
the older versions did. I thought that only affected 12.3, but udev in 12.2
might have been upgraded. The first thing to check is the rules in
/etc/udev/rules.d/ to see which ones are renaming the ethernet devices. The
first place to look is in 70-persistent-net.rules. Delete any rules that end
with “NAME=ethX”, where X is a number.

I think I’ve traced the source: https://bugzilla.novell.com/show_bug.cgi?id=753370

The problem happened when I stopped & started the machine to replace the NIC (I restarted the machine 3 times which I was setting up the bonding again, I think). I think udev “remembered” the mac addresses as persistent, but bonding was resetting the mac addresses, and over the reboot the cards “remembered” the temporary mac addresses. udev the had another look at them (on boot), and the different mac address lead to renaming, which then meant that they weren’t part of the bonding, so everything was SNAFU.

I’ve taken off bonding. I’ll delete the persistent udev entries, and we’ll see what happens …

Tried that. I now have 2 x NICs with the same MAC address! Any suggestions how I reset the MAC address?


eth0      Link encap:Ethernet  HWaddr 00:13:20:D5:AA:7B  
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:306 errors:0 dropped:1 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:52794 (51.5 Kb)  TX bytes:0 (0.0 b)
          Interrupt:18 Base address:0x2f00 

eth1      Link encap:Ethernet  HWaddr 00:13:20:D5:AA:7B  
          inet addr:192.168.xxx.yyy  Bcast:192.168.xxx.255  Mask:255.255.255.0
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:14441 errors:0 dropped:2 overruns:0 frame:0
          TX packets:35395 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:2748016 (2.6 Mb)  TX bytes:42866066 (40.8 Mb)

and /etc/udev/rules.d/70-persistent-net.rules has newly created entries:


# PCI device 0x8086:0x1064 (e100)
# PCI device 0x10ec:0x8169 (r8169)
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:13:20:d5:aa:7b", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="eth*", NAME="eth0"
SUBSYSTEM=="net", ACTION=="add", DRIVERS=="?*", ATTR{address}=="00:13:20:d5:aa:7b", ATTR{dev_id}=="0x0", ATTR{type}=="1", KERNEL=="eth*", NAME="eth1"

Any suggestions to straighten this out?

Thanks
David

On 04/07/2013 12:36 PM, jetojedno wrote:
>
> Tried that. I now have 2 x NICs with the same IP address! Any
> suggestions how I reset the MAC address?

Slight correction - they have the same MAC address. As shown below, eth0 has no
IP address.

>
> Code:
> --------------------
>
> eth0 Link encap:Ethernet HWaddr 00:13:20:D5:AA:7B
> UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> RX packets:306 errors:0 dropped:1 overruns:0 frame:0
> TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
> collisions:0 txqueuelen:1000
> RX bytes:52794 (51.5 Kb) TX bytes:0 (0.0 b)
> Interrupt:18 Base address:0x2f00
>
> eth1 Link encap:Ethernet HWaddr 00:13:20:D5:AA:7B
> inet addr:192.168.xxx.yyy Bcast:192.168.xxx.255 Mask:255.255.255.0
> UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> RX packets:14441 errors:0 dropped:2 overruns:0 frame:0
> TX packets:35395 errors:0 dropped:0 overruns:0 carrier:0
> collisions:0 txqueuelen:1000
> RX bytes:2748016 (2.6 Mb) TX bytes:42866066 (40.8 Mb)
>
> --------------------
>
>
> and /etc/udev/rules.d/70-persistent-net.rules has newly created
> entries:
>
>
> Code:
> --------------------
>
> # PCI device 0x8086:0x1064 (e100)
> # PCI device 0x10ec:0x8169 (r8169)
> SUBSYSTEM==“net”, ACTION==“add”, DRIVERS=="?", ATTR{address}==“00:13:20:d5:aa:7b”, ATTR{dev_id}==“0x0”, ATTR{type}==“1”, KERNEL=="eth", NAME=“eth0”
> SUBSYSTEM==“net”, ACTION==“add”, DRIVERS=="?", ATTR{address}==“00:13:20:d5:aa:7b”, ATTR{dev_id}==“0x0”, ATTR{type}==“1”, KERNEL=="eth", NAME=“eth1”

Do you have any special code that is changing the MAC address for one or both of
the devices? If so, that would be a problem and would explain why udev wrote two
rules for the same MAC address. By the way, the first 3 digits of the MAC
address, 00:13:20, should identify the manufacturer. According to
http://www.cavebear.com/archive/cavebear/Ethernet/vendor.html, there is no
vendor with those digits. The first value for the PCI IDs indicate that you have
an Intel device (8086) and a Realtek device (10ec).

I suggest getting rid of any bonding instructions until you get the interfaces
running in ordinary mode.

I’ve deleted all the bonding setup - it’s all back to normal;), apart from the MAC addresses.

Any way to find out the original MAC addresses, before bonding messed with them?

BTW, all was set up using YAST LAN admin, apart from deleting the udev entries.

Well, sorted both problems. I found the old MAC address in the system logs from before when I set up bonding, and set it in /etc/sysconfig/network/ifcfg-eth0. Scrubbed all the udev rules again, and it’s come up looking clean, even if it’s a cludge.

On 04/07/2013 06:16 PM, jetojedno wrote:
>
> Well, sorted both problems. I found the old MAC address in the system
> logs from before when I set up bonding, and set it in
> /etc/sysconfig/network/ifcfg-eth0. Scrubbed all the udev rules again,
> and it’s come up looking clean, even if it’s a cludge.

Yes, it is a kludge. Somewhere in your setup files, probably in the /etc/ tree,
I think you have a command that is forcing the MAC address. As I never have
tried to clone a particular MAC address, I have no suggestions on where to look.