[DRBD-user] Weird Protocol C Behaviour

[dotnetdub]

Hi All,

I'm running DRBD from Debian Squeeze (Version: 8.3.7 (api:88))

I'm having a strange issue where if I do a write on the primary I'm
getting the full speed of the BBU cached RAID card:


dd if=/dev/zero of=/test/test121212 bs=1M count=1000
1000+0 records in
1000+0 records out
1048576000 bytes (1.0 GB) copied, 1.16411 s, 901 MB/s


And the data doesn't replicate straight away. It could take upto 20
seconds for the disk on the secondary to be written to.

I'm using dedicated Gbps links with crossover cable.

Config is below.

Any ideas?

Brian


global { usage-count no; }
common { syncer { rate 35M; } }
resource r0 {
        disk {
                no-disk-barrier;
                no-disk-flushes;
        }
        protocol C;
        startup {
                wfc-timeout  15;
                degr-wfc-timeout 60;
        }
        net {
                max-buffers 8000;
                max-epoch-size 8000;
                unplug-watermark 16;
                sndbuf-size 0;
                cram-hmac-alg sha1;
        }
        on prim {
                device /dev/drbd0;
                disk /dev/cciss/c0d0p3;
                address 10.0.0.1:7788;
                meta-disk internal;
        }
        on sec {
                device /dev/drbd0;
                disk /dev/sda3;
                address 10.0.0.2:7788;
                meta-disk internal;
        }
}