Dell™ Power Edge™ 2950. Post Fail E2022

Good morning, Jamal.

The PCIe bus has failed on NIC2, causing the E2022  No POST error.  The motherboard will need to be replaced on the server to repair it.

Regards,

Hello Jamal

The PCIe bus trains to the device connected in the slot, similar to auto-negotiation of network interface speeds. Normally a reseat of the device in the PCIe slot will correct the issue. Since the error is occurring on an integrated device that is not possible.

I would recommend disconnecting all network cables from the integrated NIC. If the error persists then drain flea power and clear NVram. You can clear NVram by moving the NVram jumper to the reset position and booting up the server. Then shut the server down and move the jumper back. Disconnect power for about 1 minute to drain flea power.

If the error still persists then you can try taking the system to minimum to POST. If the error still persists then the system board will likely need to be replaced. If you are able to get into the system BIOS then you can try disabling the integrated NIC to bypass the error.

Thanks

hi,

i also had a same problem with server Dell PowerEdge 2950, on the front small lcd the error message shown like "e2022 POST fail"...and the monitor the error message shown "PCIe training error: Internal Storage slot.  System Halted..

Please guide me to solve this kind of problem, Thanks.

Ramly.

Hi Daniel, we tried the above with our 2950 to no joy.  Can you explain how you can take the system to minimum of POST?

Also is there any way to change the boot up sequence to exclude the network boot before it hits the training sequence, or disable the NIC?

much appreciate your help

Thanks Daniel - looks like I need to replace the motherboard....

Can you explain how you can take the system to minimum of POST?

There are minimum required components for the BIOS to complete POST. Minimum to POST is when you remove all unnecessary items from the system. The required items are the system board, 2 DIMMs in slots 1 and 2, 1 CPU in slot 1, both risers, the sideplane, and the control panel so you can turn on the system.

Also is there any way to change the boot up sequence to exclude the network boot before it hits the training sequence, or disable the NIC?

No, the PCI training error is a POST failure. If the system does not complete POST you will not be able to enter the BIOS.

"PCIe training error: Internal Storage slot.  System Halted..

Reseat your PERC. If the error persists after a reseat then remove the controller. If the error still persists with the PERC removed then it is an issue with the slot. Remove the sideplane the PERC is connected to. If the error persists with the sideplane removed then it is a system board issue and the system board will need to be replaced. If the error stops when removing the sideplane then the sideplane should be replaced.

PCI training errors can occur on any component that communicates on the PCI bus. If the error is occurring on an integrated device then the majority of the time the system board will need to be replaced to correct the problem. If the error occurs on a non-integrated device then reseating the component usually resolves the issue.

Thanks

Well I was just working on this same model in the INS building in buffalo, ny and on the

one I am looking at the raid controller is not in the motherboard. it is on a riser on the front

right side of the server, right were the cables run.

sorry that is the front left side of the server.