I have one filler unable to boot DOT.
After diagnose first AP was to replace the nvramm dimm but still same error (with clear mesage to rebbot and replace the motherboard)
We habe an N SERIES N6040 with single image mode dot 7.3.3 and working node has taken over faulty one.
In the doc I received:
(Two Controller modules - this is a clustered environment andconcurrent replacement is possible but not recommended)
I would like to know if it is possible to replace without maintenance window the mother board ? (without impacting production)..?
You should be fine. If the controller with the faulty motherboard has already been taken over, you can power it down, replace the motherboard, and power up without affecting anything. Depending on your environment, you may need to schedule downtime for the giveback - if you're using CIFS, for example.
I have done this once on a V3020 cluster. Usually after a boot of a taken-over controller, it comes up to "waiting for giveback" - I don't recall if, after the MB replacement, I had to do anything special to get this to happen. I think on boot it detects that the partner is in takeover mode and knows not to boot - can anyone comment?
I would advise making sure all the eeprom settings are the same on the new MB. On the old controller (prior to the replacement), drop to the eeprom level, type printenv, and capture the output. Once the new MB is installed, do the same, compare the settings, and change anything on the new MB that doesn't match the old MB (change with setenv).
Sorry, without a system in front of me I also can't recall how to get to the eeprom - ctrl-C at boot, then service mode, or special mode, or something, then bye...might be different on N-series anyway.