[DRBD-user] Serious problems with perfomance drdb

Moiseev Petr mail.moiseev at gmail.com
Wed Oct 1 09:52:35 CEST 2008

Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.


I have a system of 2 identical nodes, each of which is set raid10 of 4 SATA
drives. Synchronization is based on individual network adapters on the 1Gb
Full duplex uplink. Disk resources are published through iscsitarget. Tests
were performed using performance IOmeter.
Settings IOmeter:
maximum disk sise = 10000 sectors
starting disks sector = 10000
Access spetifications = All in one

Sample results (no errors in the accuracy of measurement):
One site works, the other is poweroff 1000 I / O, 10 Mb / s Both site work,
drbd disconnected 1000 I / O, 10 Mb / s Both site work, drbd connected 10
(!) I / O, 0.1 (!) M / s

And drbd simply in a state connected, synchronization has been completed,
the two nodes in UpToDate. What could be the problem? Details of the systems
and configuration below.

OS Ubintu 8.04.
Net 10.133.133.0/30 is simply uplink bettween nodes.

#aptitude install drbd8-utils

#  /etc/drbd.conf
global { usage-count no; }

common {             protocol A;
    startup { wfc-timeout 30; degr-wfc-timeout 120; }
    disk  { on-io-error detach; }
    net   {       timeout 60;
                  connect-int 10;
                  ping-int 10;
                sndbuf-size       240k;
                max-buffers      20480;
                max-epoch-size   16384;
                unplug-watermark 20480;
                  after-sb-0pri discard-younger-primary;
                  after-sb-1pri consensus;
          }
    syncer  {     rate 500M;
                al-extents 257;
            }
}
resource r0 {
    on fs0 {
      address 10.133.133.1:7790;
      disk /dev/sdb1;
      device /dev/drbd0;
      meta-disk "internal";
    }
    on fs1 {
      address 10.133.133.2:7790;
      disk "/dev/sdb1";
      device "/dev/drbd0";
      meta-disk "internal";
    }
}

# aptitude install iscsitarget

#  /etc/ietd.conf
Target iqn.2008-09.com:storage.r0
        IncomingUser iscsi XXXXXXXXXXXX
        Lun 0 Path=/dev/drbd0,Type=fileio

# ifconfig eth1
eth1      Link encap:Ethernet  HWaddr 00:30:48:d2:1b:37
          inet addr:10.133.133.1  Bcast:10.133.133.3  Mask:255.255.255.252
          inet6 addr: fe80::230:48ff:fed2:1b37/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:9000  Metric:1
          RX packets:31057691 errors:0 dropped:0 overruns:0 frame:0
          TX packets:66325409 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:3125652368 (2.9 GB)  TX bytes:2951365778 (2.7 GB)
          Base address:0x3000 Memory:d8400000-d8420000
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linbit.com/pipermail/drbd-user/attachments/20081001/aab228a1/attachment.htm>


More information about the drbd-user mailing list