Nasty problem with PCIe and LAN cards

Hey guys, I need help with mine big and nasty problem with my server hardware (SuperMicro H8DCi) which runs openSUSE 13.1, fully updated.
I have there 2 PCIe Intel PRO/1000 CT Desktop ethernet adapters and two onboard ethernet ports. The problem is with two ethernet PCIe LAN cards. Main board has three PCIe slots - first x16 with first LAN card, second x16 with graphics card and third x4 with second LAN card.

So, now the description of the problem:

I tried to bond together some ethernet ports. Bonding works perfect, so I ran some network copy. One hour it was great and then the problem starts. It starts to be unstable, losing packets and probably in 15 minutes its connection lost definitely. I tried to restart network service, manually specify IP address, disconnecting the cable, nothing helps. Then I tried to erase bonding and configure network on separate adapters, on the onboard ports was everything ok, but the PCIe adapters aren’t working. The two Intel adapters are visible in “hwinfo”.
I have snapshotted the system before the configuration of the bond, so I tried to reverse the changes, don’t work. I grab openSUSE DVD and reinstalled the system from fresh, nothing. Then I tried to play with the cards itself, I changed one LAN card with another in ports, nothing again. so I changed one adapter with the graphics card and then the one card starts working, but the graphics card starts to be weird, the GUI won’t start on the card, but in the another PCIe slot it works perfect.

The problem adapters are eth0 and eth1

So, I don’t know what happened at the main board.
I will be very pleased if you can help resolve my problem. I’m very sorry for the long thread.

What outputs you will need for helping me out?

I have the output from ifconfig:

eth0 Link encap:Ethernet HWaddr 68:05:CA:1B:09:A8
UP BROADCAST MULTICAST MTU:1500 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:0 (0.0 b) TX bytes:0 (0.0 b)
Interrupt:43 Memory:dfde0000-dfe00000

eth1 Link encap:Ethernet HWaddr 68:05:CA:1C:84:9E
UP BROADCAST MULTICAST MTU:1500 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:0 (0.0 b) TX bytes:0 (0.0 b)
Interrupt:42 Memory:dfbe0000-dfc00000

eth2 Link encap:Ethernet HWaddr 00:30:48:59:F8:0A
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:1623 errors:1 dropped:0 overruns:0 frame:1
TX packets:1193 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:2108598 (2.0 Mb) TX bytes:120562 (117.7 Kb)

eth3 Link encap:Ethernet HWaddr 00:30:48:59:F8:0B
inet addr:10.0.0.70 Bcast:10.0.0.255 Mask:255.255.255.0
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:15675 errors:0 dropped:0 overruns:0 frame:0
TX packets:6034 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:8751167 (8.3 Mb) TX bytes:708744 (692.1 Kb)

lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
UP LOOPBACK RUNNING MTU:65536 Metric:1
RX packets:232 errors:0 dropped:0 overruns:0 frame:0
TX packets:232 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:20759 (20.2 Kb) TX bytes:20759 (20.2 Kb)

ok this may be outdated info but I always put the video in the first slot. There was a good reason for this long ago not sure it still applies

Also read the release notes there was a change in the naming of ports

Thanks for responding gogalthorp

this main board is with two processor slots and first marked slot is the second x16 in a row. When you have only one processor installed, then only this slot one, but seconds in row is working. So I mean to have graphics card installed correctly.

I got something. This could cause problems due to faulty CPU, even when working correctly? Because I can not see without a second processor 4 memory banks, one PCIe x16 slot, one PCIe x4 slot and one onboard ethernet port.