[DRBD-user] Question about WAN replication using 8.3.10

[Sean McCreadie]

Hello,

I just setup two test DRBD boxes connecting over my WAN running drbd 8.3.10. I am seeing poor write performance during testing and would appreciate any help.

I am using protocol A.  My WAN circuit is capable of sustained 35+ Mb/s throughput and about 12ms RTT latency.  I was hoping this would be enough for decent primary performance.

I understand that writes will be slowed down on the primary node until the send buffer is cleared during heavy writes, so I would like any help with this setting.  I am also interested in the ahead/behind mode but am unsure how to enable it properly.  I have included my recent drbd.conf, I am testing with it now.  If anyone has any tips, thank you in advance.



global {
    usage-count ask;
}

common {
  syncer { rate 10M; }
}

resource vg0_drbd {

  protocol A;

  handlers {
    pri-on-incon-degr "echo O > /proc/sysrq-trigger ; halt -f";
    pri-lost-after-sb "echo O > /proc/sysrq-trigger ; halt -f";
    local-io-error "echo O > /proc/sysrq-trigger ; halt -f";

  }

startup {
   degr-wfc-timeout 120;    # 2 minutes.
  }

disk {
    on-io-error detach;
    no-disk-barrier;
    no-disk-flushes;
    no-md-flushes;

  }

net {
     max-buffers 2048;
     max-epoch-size 2048;
     unplug-watermark 2048;
     sndbuf-size 0;

  }

  syncer {
  al-extents 3389;
  }


  on node01 {
    device     /dev/drbd0;
    disk       /dev/sdb;
    address    172.18.10.50:7788;
    meta-disk internal;
}

   on node02 {
    device    /dev/drbd0;
    disk      /dev/sdb;
    address   172.19.10.50:7788;
    meta-disk internal;
  }
}

Thanks,

Sean



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linbit.com/pipermail/drbd-user/attachments/20110325/2643bc26/attachment.htm>