Sudden infinite reboot on PXE boot

  • Hi to the community. Today, 5 of my 400 machines couln’t boot from PXE.

    1. they pass the DCHP call
    2. then I get the “configuring net ( macadress )” step
    3. then they reboot and get to the 1.

    The strange thing is that I don’t make changes on the FOG server (1.2.0) or on the machines. And they booted perfectly til today. ( and no other machine has this issue ).

    they are all Lenovo M70 ( but 1 other don’t have this problem )

    Is anyone of you have an idea …

  • So here is just a feedback : you were totally right !
    The problem causes now very low speed while imaging on some machines ( record : 22 MB/min !!! ) and as soon as I connect the machine on another switch, the speed increase frankly. I’m not victorious because I didn’t isolate the faulty swith ( there are many interconnected … ) even if I suspect one ( but reboot it doesn’t change anything.
    So I don’t bother you and ask more as it is clearly a network issue and not a FOG one, but when I found a solution ( I hope so … )
    I will post it here …

    Thank you again !

  • I think they’re on different switches, but I’ll check and reboot them tomorrow.
    Thanks a lot for your fast and wise support

  • Senior Developer

    I’d have to agree with Jaymes on this one.

    Are the systems having the issue on the same switch? Is there a broadcast storm potentially happening on the scope of that network?

    It seems odd, to me, that all was working, then poof, they just stop working. If nothing changed server wise, and about 99% of the other systems are still functioning, I’d start trying to see what the “commonalities” are.

  • Developer

    Did you try to reboot the switches? Did you try a different port ( I pull the culprit from it’s environment and test in my office to be sure). Your fog server seems to be functioning as expected, considering most of your machines are booting correctly.

    I have seen this issue in a few of my labs, usually rebooting the switches (managed or unmanaged) will solve this issue. This issue sounds network related.