iLO ARP cache

[juanmarco]

iLO 6 v1.59p22 on ProLiant DL360 Gen11 Server fails to update the MAC address for its default gateway.

It seems to ignore gratuitous ARPs from the replaced gateway device. Also it doesn't seem to expire its ARP cache as we have confirmed by port mirror that it's still trying to send packets via old MAC address of default gateway, over a week after gateway migration.

Support isn't helpful as usual "There should not be any issue ILO6 as ot was working fine before firewall replacement." (sic)

Any idea how to remedy short of manually somehow getting to local LAN, and manually reseting all iLOs? Already tried bouncing the management ports (port down/up) on the switches, iLO remains unimpressed.

[shiva_jr]

Hi @juanmarco , 
   Will this Advisory helps you ?  Please let me know if still the issue persists.
Regards,
Shiva_jr
Give 'Kudos' and mark as "Accepted solution' if my post solve your issue.

[juanmarco]

That is old stuff. iLO5 / 2019

But there seem to have always been issues with this, so it's probable they have persisted across the releases.

[ngnear]

I'd suggest to get the firmware version updated to the latest and check if the issue prevails.

[juanmarco]

The release notes don't say anything about this.
Also it will be difficult to reproduce as this is more or less a one-time incident and as soon as we restart the iLOs (even in the current version) the problem is gone anyway.

We usually only apply SPPs.

[ngnear]

iLO firmware update is a non disruptive in nature. Your production will not impacted.

If you face the same error again, feel free to update the latets iteration.

[juanmarco]

@ngnear We do regular updates.

1) Newer versions do not mention this problem in the release notes anyway.

2) Once we reboot the issue will be gone anyway (as

I explained) and we don't replace routers every day (which was the initial issue, in case anyone read the problem description).

[ngnear]

Since the issue is intermittent, we might need to capture and check some logs.
This also might require us to try and replicate the issue in our labs.
We suggest you to log a support case for further extended diagnosis on this at support.hpe.com