Unsolved
This post is more than 5 years old
2 Posts
0
63568
poweredge 2900 degraded link width error
Hi all,
we have here a poweredge 2900 server which halts during the bios checks after a reboot
with the following error:
degraded link width error: embedded i/o bridge device 1
expected link width is x4
actual link width is x1
There is only one PCIe card (video) sitting in the slots. I've done the following tests:
i replugged the video card, disconnect power cable and wait some time, but nothing helps.
Every hint is welcome
Thanks in advance
Ollie
Dev Mgr
9.3K Posts
0
June 10th, 2010 12:00
That error is most likely related to the PERC5i or PERC6i that's in the server. These use a x4 SAS channel (4 x 300MByte/s) for bandwidth and failover/redundancy. It sounds like something is wrong with the PERC, the cable, or the backplane.
You may want to open a case with Dell PowerEdge support. To be a step ahead of them, you could run a DSET on the server (support.dell.com/dset) for them to look at.
ollie-red
2 Posts
0
June 10th, 2010 23:00
i've also taken the PERCI6i interface card out, but booting stops at the same test. Resetting the system to the default values helps not. So i believe there is no chance to run any operation system.
Does anyone has more ideas?
ricepower
6 Posts
0
June 15th, 2010 08:00
I do encounter this link width error problem with our HPC clusters. Could you pls look over on the bmc sel log which device id is causing the link width error? . I suspect it's either the internal nic or LOM, or the chipset.
Anyhow, the only solution is to replace the mainboard.
CalebHart
1 Message
0
February 25th, 2013 02:00
Hi all,
we have here a poweredge 2950 server which halts during the bios checks after a reboot
with the following error:
degraded link width error: embedded i/o bridge device 18
expected link width is x8
actual link width is x4
system halted!