[c-nsp] POST PortASIC Failure: Fail PortPhyLoopback

[Andy Ellsworth]

On Wed, Jan 16, 2013 at 2:39 PM, Williams, Ray <ray.williams at transcore.com>
 wrote:

And, this is what the POST message looks like when the port *does not* connect
> to the DEMARC after a reboot.
>
> POST: PortASIC Port Loopback Tests : Begin****
>
> POST: Failed PortPhyLoopback Packet Test asic_index 1 port_hardware_index
> 26****
>
> POST: PortASIC Port Loopback Tests : End, Status Passed****
>
>
This is not really a solution, just an anecdote to confirm that you're not
entirely crazy to infer that attached devices can cause squirrely POST
results.

Several years ago, we had Catalyst 2950 switches connected to the
integrated Gigabit switches in Dell PowerEdge 1655MC blade chassis. If the
blade chassis and switch were powered up simultaneously (a common
occurrence when recovering from frequent power/UPS/generator failures in a
third world call center), the ports on the 2950 that connected to the
1655MC would invariably fail the loopback tests during POST and be disabled
until the 2950 was power cycled, at which point the ports on the 2950 would
come up fine. My conclusion was that something about the state of the
Ethernet interfaces in the 1655MC during *its* bootup phase caused the 2950
loopback POST sequence to freak out. Once the blade switch was fully
booted, the 2950 POST sequence would pass without incident.

-Andy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://puck.nether.net/pipermail/cisco-nsp/attachments/20130501/4a3cd8ef/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.gif
Type: image/gif
Size: 530 bytes
Desc: not available
URL: <https://puck.nether.net/pipermail/cisco-nsp/attachments/20130501/4a3cd8ef/attachment-0001.gif>