drbd-user May 2010 archive
Main Archive Page > Month Archives  > drbd-user archives
drbd-user: [DRBD-user] DRBD performance - disk random access tim

[DRBD-user] DRBD performance - disk random access time problem

From: Proskurin Kirill <proskurin-kv_at_nospam>
Date: Mon May 31 2010 - 12:09:53 GMT
To: "drbd-user@lists.linbit.com" <drbd-user@lists.linbit.com>

Hello.

Software:
Debian Lenny 2.6.26-2-amd64
ocfs2-tools 1.4.1-1
drbd8-2.6.26-2-amd64 2:8.3.7-1~bpo50+1+2.6.26-21lenny4
drbd8-source 2:8.3.7-1~bpo50+1
drbd8-utils 2:8.3.7-1~bpo50+1

Hardware:
Dell server
Adaptec 5405 RAID Controller (write cache enabled)
2x1TB 7200rpm HDD

I make some test for disk random access time and it is horrible:

# ./seeker /dev/drbd0
Seeker v2.0, 2007-01-15,
http://www.linuxinsight.com/how_fast_is_your_disk.html
Benchmarking /dev/drbd0 [858643MB], wait 30
seconds..............................
Results: 65 seeks/second, 15.26 ms random access time

While linear speed is good.
# hdparm -Tt /dev/drbd0

/dev/drbd0:
  Timing cached reads: 13928 MB in 2.00 seconds = 6972.83 MB/sec
  Timing buffered disk reads: 628 MB in 3.01 seconds = 208.82 MB/sec

Whats wrong with seek time? How can I tune\fix it?

Seeks on non DRBD partition is more good:

Results: 129 seeks/second, 7.71 ms random access time

My drbd conf:

# cat /etc/drbd.d/global_common.conf
global {
        usage-count yes;
}

common {
        protocol C;

        handlers {
                pri-on-incon-degr "/usr/lib/drbd/notify-pri-on-incon-degr.sh;
/usr/lib/drbd/notify-emergency-reboot.sh; echo b > /proc/sysrq-trigger ;
reboot -f";
                pri-lost-after-sb "/usr/lib/drbd/notify-pri-lost-after-sb.sh;
/usr/lib/drbd/notify-emergency-reboot.sh; echo b > /proc/sysrq-trigger ;
reboot -f";
                local-io-error "/usr/lib/drbd/notify-io-error.sh;
/usr/lib/drbd/notify-emergency-shutdown.sh; echo o > /proc/sysrq-trigger
; halt -f";
                outdate-peer "/usr/lib/heartbeat/drbd-peer-outdater -t 5";
                        }

        startup {
                wfc-timeout 60;
                degr-wfc-timeout 30;
                outdated-wfc-timeout 15;
                become-primary-on both;
                }

        disk {
                fencing resource-and-stonith;
                no-disk-flushes;
                no-md-flushes;
                }

        net {
                cram-hmac-alg sha1;
                shared-secret "password";
                allow-two-primaries;
                ping-timeout 20;
                 after-sb-0pri discard-zero-changes;
                 after-sb-1pri discard-secondary;
                 after-sb-2pri disconnect;
                }

        syncer {
                rate 100M;
                }
}

# cat /etc/drbd.d/drbd0.res
resource drbd0 {

on server1.domain.org {
device /dev/drbd0;
disk /dev/sda9;
address 192.168.1.1:7789;
meta-disk internal;
}

on server2.domain.org {
device /dev/drbd0;
disk /dev/sda9;
address 192.168.1.2:7789;
meta-disk internal;
}

  }

-- С уважением, Проскурин Кирилл _______________________________________________ drbd-user mailing list drbd-user@lists.linbit.com http://lists.linbit.com/mailman/listinfo/drbd-user