ACI - Port Configuartion Failure - Fault F1186

[ju.mahieu]

Hi,

I have the following error:

"Port configuration failure. Reason: 2 Failed Config: l1:PhysIfautoNeg_failed_flag" on my "topology/pod-1/node-101/sys/phys-[eth152/1/13]"

Explanation: This fault is caused by a hardware programming failure
Recommended Action: The system will periodically retry to program the hardware. If the failure is due to lack of hardware resources, free up resources by removing or simplifying existing configuration.

How can I check my system resources or solve this issue ?

Thank you.

I wish you an happy new year ;-)

Ju

[Nik Noltenius]

Hi Ju,

if you haven't already done that, I would focus on the "autoNeg_failed_flag" part of the fault description. What do you have connected to the port and is it possible that speed and duplex auto negotiation does not work properly?

Does the Link Level Policy (Fabric > Access Policies > Interface Policies > Link Level) that you've attached to the interface allow auto negotiation?

Regards,
Nik

[ju.mahieu]

Hi Nik,

Thank you for your answer.

Unfortunately, the interface isn't anymore attached to any policy (picture attached).

Regards

Ju

[Nik Noltenius]

Hi Ju,

that's strange, indeed.

Unfortunately I do not have any FEX connected to my fabric so this is more like poking around in the dark:

Is that fault the only one you get or are there more possibly related to this one?

Does the output from your screenshot look normal compared to other FEX ports or is it different? The reason I'm asking is that for a server facing port on my leaf switches I get a whole list of properties and they are used in a couple of policies even if they have never been configured and are operationally down (see picture attached). Thus your screenshot looks rather strange to me but anyway that could be FEX-specific.

Have you tried reconnecting the port to some end-system and if so, does the fault disappear then?

Do you have hosts attached to other FEX interfaces and if so, can you disconnect one for testing purposes and figure out if that port shows the same fault?

On the leaf switch there is a "show system internal fex" command providing a couple of sub-commands. Maybe that provides some intel if you haven't already tried that.

Regards,

Nik

[ju.mahieu]

Hi Nik,

The problem has been resolved after a fex boot.

Regards,

Ju

=============================

Below the answers to your questions :

=============================

Is that fault the only one you get or are there more possibly related to this one?
** I get only one issue of this kind.  

Does the output from your screenshot look normal compared to other FEX ports or is it different?
** Same screenshot on all FEX ports

The reason I'm asking is that for a server facing port on my leaf switches I get a whole list of properties ...
** Right. For the same port, there is different output if you use :
- /aci/fabric/inventory/pod-1/fb-dij-leaf101/fabric-extenders/152/extended-chassis-modules/1/host-ports/13
# extended-chassis-host-port
id             : 13
...
type           : extended-chassis-host-port

or

- /aci/fabric/inventory/pod-1/fb-dij-leaf101/interfaces/physical-interfaces/eth152:1:13
interface            : eth152/1/13
admin-state          : up
usage                : discovery
bandwidth            : 0
delay                : 1
mdix                 : auto
medium               : broadcast
...
eee-lat              : variable
eee-lpi              : aggressive
eee-state            : not-applicable
portcap-speed        : 100,1000,auto
channeling-state     : unknown

Do you have hosts attached to other FEX interfaces and if so, can you disconnect one for testing purposes and figure out if that port shows the same fault?
*** For now , I don't have hosts connected to my Fex.

On the leaf switch there is a "show system internal fex" command providing a couple of sub-commands. Maybe that provides some intel if you haven't already tried that.
*** All output from "show system internal fex" seems to be normal.