[DRBD-user] drbd-xen and fencing

[Vadym Chepkov]

Hi,

I am using drbd-xen helper script for XEN live migration and I can't figure out how to properly fence the VM in pacemaker:

I tried this way:

common {
  protocol C;
  startup {
    wfc-timeout 15;
  }
  syncer { 
    rate 100M;
    verify-alg sha1;
  }
  disk {
    on-io-error detach;
    fencing resource-only;
  }
  net {
    after-sb-0pri discard-zero-changes;
    after-sb-1pri discard-secondary;
  }
  handlers {
        split-brain "/usr/lib/drbd/notify-split-brain.sh root";
        out-of-sync "/usr/lib/drbd/notify-out-of-sync.sh root";
  }
}

….

resource vsvn {
  net {
    allow-two-primaries;
  }
  handlers {
    fence-peer "/usr/lib/drbd/crm-fence-peer.sh -i xen_vsvn";
    after-resync-target "/usr/lib/drbd/crm-unfence-peer.sh -i xen_vsvn"; 
  }
  device    /dev/drbd4;
  disk      /dev/vg2/vsvn;
  flexible-meta-disk internal;
  
  on xen-11 {
    address   10.0.0.1:7785;
  }
  on xen-12 {
    address   10.0.0.2:7785;
  }
}

But, when one of the nodes going down the constraint is not getting added. How can I fix it?

# cat /proc/drbd 
version: 8.3.9 (api:88/proto:86-95)
GIT-hash: 1c3b2f71137171c1236b497969734da43b5bec90 build by mockbuild at centos64-dev.chepkov.lan, 2010-10-27 18:50:37

Thank you,
Vadym