[DRBD-user] system freezes with drbd 0.7.21 on kernel 2.6.18

[Joachim Rieger]

Hello Florian,

ok, I updated drbd yesterday to version 8.2.4 but the problem still exists.
(my test: dd if=/dev/zero of=dd-test1 bs=1024k count=2000 on Xen container
and after that, the Xen Container and the Dom0 just hangs.)
Problem also appears if heartbeat is down.
Problem does not appear when one host is down, and both drbd-partitions
are active on one host.
Can it be something with the NIC ?
drbd and heartbeat are connected through 1Gb cross connection.

my drbd.conf ( some things are commented out, because of the upgrade or
testing)

resource drbd0 {
  protocol      C;
# incon-degr-cmd "echo '!DRBD! pri on incon-degr' | wall ; sleep 60 ;
halt -f";
  handlers { pri-on-incon-degr "logger '!DRBD!'; logger `cat /proc/drbd`"; }
  startup { wfc-timeout 0; degr-wfc-timeout     120; }
  disk { on-io-error detach; }
  syncer { rate 100M; }
#  syncer {
#    rate        50M;
#    group       1;   #
#  }

  on rod {
    device      /dev/drbd0;
    disk        /dev/md2;
    address     10.0.0.1:7791;
    meta-disk   internal;
  }
  on todd {
    device      /dev/drbd0;
    disk        /dev/md2;
    address     10.0.0.2:7791;
    meta-disk   internal;
  }
}

resource drbd1 {
  protocol      C;
# incon-degr-cmd "echo '!DRBD! pri on incon-degr' | wall ; sleep 60 ;
halt -f";
#  incon-degr-cmd "logger '!DRBD!'; logger `cat /proc/drbd`";
  handlers { pri-on-incon-degr "logger '!DRBD!'; logger `cat /proc/drbd`"; }
  startup { wfc-timeout 0; degr-wfc-timeout     120; }
  disk { on-io-error detach; }
  syncer { rate 100M; }
#  syncer {
#    rate        50M;
#    group       2;   #
#  }

  on rod {
    device      /dev/drbd1;
    disk        /dev/md3;
    address     10.0.0.1:7792;
    meta-disk   internal;
  }
  on todd {
    device      /dev/drbd1;
    disk        /dev/md3;
    address     10.0.0.2:7792;
    meta-disk   internal;
  }
}

Xen domU config:
/etc/xen/openex.cfg:

#
kernel = "/boot/vmlinuz-2.6.18-4-xen-amd64"
ramdisk = "/boot/initrd.img-2.6.18-4-xen-amd64"
name="openex"
memory=1024
disk=['file:/home/xen2/openex.img,hda1,w','file:/home/xen2/openex_swap.img,hda2,w']
root = "/dev/hda1"
# network
vif=[ 'bridge=xenbr0' ]
dhcp="off"
hostname="openex"
extra = "3"

Thanks!
Joachim



> Hello,
>
> 1. 0.7 is outdated;
> 2. You are 4 releases behind even the current 0.7 release;
> 3. We can't guess your drbd.conf, or your Xen domU config for that matter;
> 4. Upgrade to at least 0.7.25 to see if your issue persists;
> 5. If it does, please post again and include relevant configuration files.
>
> And btw Heartbeat 2.0.7 is outdated (and buggy too), even though that is 
> unlikely to be the root cause of your problem.
>
> Florian
>
> On Monday 28 January 2008 13:19:54 Joachim Rieger wrote:
>   
>> Hi,
>>
>> I have the following setup:
>>
>> 2 Hosts with Debian Etch and Kernel 2.6.18-4  ( 2.6.18-4-xen-amd64).
>> On a SoftRAID/md (RAID1) I have two drbd partitions (drbd 0.7.21) with XEN
>> (3.1) Containers on it and running heartbeat(2.0.7)
>>
>> When I produce disk i/o on the Xen Containers (e.g. dd 2gb to a file or cp
>> big file ) or on the drbd FS the whole system hangs or freezes
>> without writing any logs and I have to 'hard reset' the machine.
>>
>> It just happens on the drbd devices.
>> Is it a bug, should I update to drbd 8.x.y ??
>> Is it the 'drbd 2.6.22 hang' Problem ?
>> Do I have to update drbd to 8.0.6 or higher ?
>>
>> Thanks in advance!
>> Joachim
>>
>>
>>
>>
>> _______________________________________________
>> drbd-user mailing list
>> drbd-user at lists.linbit.com
>> http://lists.linbit.com/mailman/listinfo/drbd-user
>>     
>
>
>
>   


-- 
--- 
Joachim Rieger

agimatec GmbH
Adenauerallee 73
D-53113 Bonn

phone   +49(0)228 184 869 80 - 31
telefax +49(0)228 184 869 80 - 15
e-mail: joachim.rieger at agimatec.de
URL: http://www.agimatec.de

Geschäftsführer: Jürgen Lüders
HRB 14607 Amtsgericht Bonn