dom0 rebooting without any reason

Discussion in 'Kernel Questions' started by divine, Nov 30, 2009.

  1. divine

    divine New Member

    Hello,

    I have some kind of problem with my xen server.
    Here is some basic info:
    cat /proc/version
    Linux version 2.6.26-2-xen-686 (Debian 2.6.26-17lenny2) ([email protected])
    4gb ram
    xeon 2ghz

    two domus: with 2048 and 256 of memory.


    Now, every two hours or so dom0 reboots without any reason, no log messages, nothing. If i stop xend (xend stop) then server works without any reboots perfectly. Please help me to identify problem, i know it is difficult because i have no information what can cause it. Hardware was tested throughtly for two weeks - everything is perfect.
    It works without any gliches with xend started too except rebooting 'thing'.
    Is there any way to 'catch dump' when that happens? (unfortunately kernel has no errors too)

    Sincerely,
    me
     
  2. falko

    falko Super Moderator Howtoforge Staff

    Did you check the logs on the host and the guests?
     
  3. divine

    divine New Member

    Yes.

    Logs capture nothing...
    So what i have got:
    1. xend stop = no reboots everything is working
    2. xend start = hangs every 2-3 hours randomly
    3. xend start but with no domus running = hadn't crashed yet (about one full day).

    my guess is RAM but i changed them, i don't know even how many times...
     
  4. falko

    falko Super Moderator Howtoforge Staff

    That's really strange. Unfortunately I have no idea what could cause this... :(
     
  5. glycerin

    glycerin New Member

    dom0 rebooting unexpectedly

    Hi ..I have the same problem. The dom0 restart without any reason and no log available. It seems that all is ok but I can understand why these reboots random.
    Did you find any solution?
    I've updated the kernel release but nothing to do. All is strange and also I cannot release the VM for production use.
    Any help?
    Thanks glycerin
     
  6. divine

    divine New Member

    Hi there,

    I researched this problem for a whole year and never found anything.
    But i have my own conclusions anyway.
    What i did:
    1. changed ram etc..
    2. tested everything with everything possible
    3. observations:
    a) hanging occurs when clustered servers are put together
    b) though it is very little possibility, but my conclusion is that somehow xen kernel+ glusterfs (sync) played a big role in this hanging. I played with various kernels and got to the point when only one server hangs constantly but other is ok.

    (now i just have drbd+xen, running without problems atm)

    Could you please write all the software you are using?
     
  7. glycerin

    glycerin New Member

    Hi. Actually I'm using simpply the hypervisor only with 4 VM but only for testing.
    Now we decided to reinstall the server because at first attempt I've compiled Xen and the xen-tools. In a second step I've decided to use the debian package and may be this approach has done something wrong in the hypervisor. These are only some hint to understand what's happening.
    Thannks
     

Share This Page