UNSOLVED

joergriether

updated

14 years ago

J

joergriether

127 Posts

0

13314

February 28th, 2013 06:00

DELL R710 system bios update takes down network for about 6 seconds

Hello,

here comes a really weird problem we see with certain r710s (where esxi is installed) after a system bios update. when we update the system bios of one of the r710s (booting with the cdu 2.1 and after that with the suu 7.2.0), suddenly after the auto power-off-on, we lose all the network connectivity in the whole area (our testlab in this case) where the server is located for about 6 seconds.

After the 6 seconds, everything is up and running fine again. But during these seconds: No pings, no networking at all. this concerns all systems connected to the switch cascades in our testlab (many 6248s, many 8024s).

our thoughts regarding this: the r710s have two 10 gig nics and 8 1gig nics connected to 8024s (the 10 gigs) and 6248s (the 1gigs). could it be that somehow the 6248s or the 8024s see a complete takedown of a huge number of interfaces at once and thus deciding to initiate a global spanning tree topology recalculation? as the switches are configured to rstp the roundabout 6 seconds would make sense in our opinion. but this is just a first guess. Do you have any ideas or had similar experience?

Best regards
Joerg

  • DELL-Chris H

    7 Practitioner

    9682 Posts

    48046 Points

    134

    0

    Posted February 28th, 2013 08:00

    Joe,

    The BIOS update will cause the system to reinitialize with the new BIOS to take effect, that is probably the lose of network you are referring to. Also, what is the role of this system in your network?

    DELL-Chris H

    Social Media and Communities Professional

    Dell Technologies | Enterprise Support Services

    #IWork4Dell

    Did I answer your query? Please click on ‘Mark as Accepted Answer’. ‘Thumbs up’ the posts you like!

  • joergriether

    127 Posts

    134

    0

    Posted February 28th, 2013 09:00

    Hi Chris,

    it´s a simple ESXi machine, just equipped with tons of nics. Thus, my guess is that the huge amount of nics going down during the post bios upgrade phase could cause the switches to somehow "soft-panic" and trigger a global stp recalculation.

    But then again, just guessing. And since (as far as i know - maybe you can set that straight) the 6248s and 8024s log files don´t differentiate between standard topology changes (which occur a thousand times a day because every system start / shutdown is a topology change) and special topology changes causing the whole paths to be recalculated - it would be very hard to locate the real causer when checking the switch logs. am i wrong? what do you think?

    Best regards,

    Joerg