Hi all,
I always used HP and Aruba switches in our system (1820 or 1920 for simple applications, 2930F for the more complex and high demand ones).
Now we force to move to Aruba 6100 (unfortunately 6200 that better fits our needs are not available in reasonable time)
Time to time, i find differences in the behavior of these new switches respect to the ones that we used to use.
Sometimes this behavior is really strange and really hard to understand and debug (if someone wants I will explain the strange behavior with multicast if the multicast destination IP is not set correctly, that i thought it is a nasty bug and not a feature)
Anyway, the problem of this week is this one:
The bootstrap firmware for one of our devices has a little bug (endianness problem) and send UDP telegrams where the two bytes of the IPv4 CRC are inverted. The application firmware instead does not have this bug.
The Aruba 6100 drops these telegrams and then the devices is not accessible during the bootstrap phase.
All the other switches that I test (also from other producers) does not drops these telegrams and then we never notice the problem.
Is there any way to temporary disable this behavior?
My problem is that the bootstrap firmware, for safety reason, can be updated only if the device is stopped during the bootstrap phase and I'm not be able to stop it to fix update the firmware that fix the bug (a kind of chicken/egg problem).
I "read" almost all the CLI documentation but I can't find anything that let me to configure this behavior.
Thanks for any suggestion