[DRBD-user] Strange HDD Performance problems with DRBD and DRBD-Proxy between two Citirx XenServern

ds at ican-gmbh.de ds at ican-gmbh.de
Fri Mar 12 17:00:06 CET 2010

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


Hello,

 

we use two Citrix XenServer 5.5 with DRBD and DRBD-Proxy and we have slow HDD performance.

 

We use an extra NIC for DRBD on both servern and the connection for DRBD is a SDSL with 2MBit sym.

For dom0 is dom0_mem configured with 2048MB.

 

Here is our /etc/drbd.conf:

 

global { usage-count yes; }

common {

  protocol A;

  net {

    shared-secret "xenserver";

    after-sb-0pri discard-zero-changes;

    after-sb-1pri consensus;

    after-sb-2pri disconnect;

    ping-timeout 20;

  }

 

  disk {

    max-bio-bvecs 1;

  }

 

  startup {

    # leave this disabled for now

    become-primary-on bornheim;

  }

 

  handlers {

    split-brain "/usr/lib/drbd/notify-split-brain.sh";

  }

 

  syncer { rate 284k; }

}

 

resource drbd-sr {

  device /dev/drbd1;

  disk /dev/sda3;

  meta-disk internal;

 

  syncer {

    csums-alg md5;

  }

 

  proxy {

    compression on;

    memlimit 1024M;

  }

 

  on bornheim {

    address 127.0.0.1:7789;

    proxy on bornheim {

      inside 127.0.0.1:7788;

      outside 192.168.93.1:7788;

    }

  }

 

  on koeln {

    address 127.0.0.1:7789;

    proxy on koeln {

      inside 127.0.0.1:7788;

      outside 192.168.94.1:7788;

    }

  }

}

 

 

One VM is Debian Lenny with bonnie++.

 

Some tests with bonnie show this:

bonnie with drbd and primary/secondary uptodate

 

Version  1.03       ------Sequential Output------ --Sequential Input- --Random-

                    -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--

Machine        Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP

test           512M 13325  22 12646   1   275   0 46463  34 129431   1   2.4   0

                    ------Sequential Create------ --------Random Create--------

                    -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete--

              files  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP

                 16 +++++ +++ +++++ +++ +++++ +++ +++++ +++ +++++ +++ +++++ +++

test,512M,13325,22,12646,1,275,0,46463,34,129431,1,2.4,0,16,+++++,+++,+++++,+++,+++++,+++,+++++,+++,+++++,+++,+++++,+++

 

 

The VM's (Windows Server 2000) are very slow and hang sometimes.

 

We observed that the primary HDD's sometimes in use (flashing) and the secondary HDD's permanently in use. And there is much traffic on the DRBD network.

 

I don't know where the error is ... Any ideas ?

 

Thanks 

 

 

Mit freundlichen Grüßen

Dietmar Simons

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

 ican GmbH
     Wanheimer Str. 408
     47055 Duisburg
     Tel.:  +49 (0) 203 73804 249
     Fax:  +49 (0) 203 73804 300
     ds at ican-gmbh.de <mailto:dietmar.simons at ican-gmbh.de> 

     http://www.ican-gmbh.de <http://www.ican-gmbh.de/> 

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

      Geschäftsführer: Khodayar Alambeigi

      Registergericht: Amtsgericht Duisburg, HRB9504

      USt-IdNr. DE 202 074 039
      Steuer-Nr. 29/457/5113

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.linbit.com/pipermail/drbd-user/attachments/20100312/8a149384/attachment.htm>


More information about the drbd-user mailing list