<html><head></head><body><div style="font-family: Verdana;font-size: 12.0px;"><div>Hi,</div>
<div> </div>
<div>i'm currently testing DRBD and i'm have read / write problems.</div>
<div> </div>
<div>On the Local Raid 5 array with BBU i can read / write with around 800 MB/s.</div>
<div>If i use /dev/drbd100 it drops to 200 MB/s</div>
<div> </div>
<div>
<div>resource drbd.test {<br/>
protocol C;</div>
<div> volume 0 {<br/>
device /dev/drbd100;<br/>
disk /dev/vgKVM.OS/drbdtest;<br/>
meta-disk internal;<br/>
}</div>
<div> net {<br/>
shared-secret "drbd.drbdtest";<br/>
}</div>
<div> on KVMe001 {<br/>
address 172.16.0.1:7100;<br/>
}</div>
<div> on KVMe002 {<br/>
address 172.16.0.2:7100;<br/>
}<br/>
}<br/>
</div>
<div>
<div># DRBD is the result of over a decade of development by LINBIT.<br/>
# In case you need professional services for DRBD or have<br/>
# feature requests visit http://www.linbit.com</div>
<div>global {<br/>
usage-count no;<br/>
# Decide what kind of udev symlinks you want for "implicit" volumes<br/>
# (those without explicit volume <vnr> {} block, implied vnr=0):<br/>
# /dev/drbd/by-resource/<resource>/<vnr> (explicit volumes)<br/>
# /dev/drbd/by-resource/<resource> (default for implict)<br/>
udev-always-use-vnr; # treat implicit the same as explicit volumes<br/>
# minor-count dialog-refresh disable-ip-verification<br/>
# cmd-timeout-short 5; cmd-timeout-medium 121; cmd-timeout-long 600;<br/>
}</div>
<div>common {<br/>
handlers {<br/>
#fence-peer "/usr/lib/drbd/crm-fence-peer.sh";<br/>
#after-resync-target "/usr/lib/drbd/crm-unfence-peer.sh";<br/>
}</div>
<div> startup {<br/>
# wfc-timeout degr-wfc-timeout outdated-wfc-timeout wait-after-sb<br/>
wfc-timeout 30;<br/>
degr-wfc-timeout 120;<br/>
outdated-wfc-timeout 30;<br/>
}</div>
<div> syncer {<br/>
rate 300M;<br/>
verify-alg md5;<br/>
}</div>
<div> options {<br/>
cpu-mask 2;<br/>
}</div>
<div> disk {<br/>
on-io-error detach;<br/>
disk-flushes no;<br/>
disk-barrier no;<br/>
c-plan-ahead 0;<br/>
c-fill-target 24M;<br/>
c-min-rate 80M;<br/>
c-max-rate 720M;<br/>
al-extents 6007;<br/>
fencing resource-only;<br/>
}</div>
<div> net {<br/>
# protocol timeout max-epoch-size max-buffers<br/>
# connect-int ping-int sndbuf-size rcvbuf-size ko-count<br/>
# allow-two-primaries cram-hmac-alg shared-secret after-sb-0pri<br/>
# after-sb-1pri after-sb-2pri always-asbp rr-conflict<br/>
# ping-timeout data-integrity-alg tcp-cork on-congestion<br/>
# congestion-fill congestion-extents csums-alg verify-alg<br/>
# use-rle<br/>
allow-two-primaries no;<br/>
after-sb-0pri discard-zero-changes;<br/>
after-sb-1pri discard-secondary;<br/>
after-sb-2pri disconnect;<br/>
rr-conflict disconnect;<br/>
max-buffers 8000;<br/>
max-epoch-size 8000;<br/>
sndbuf-size 2M;<br/>
rcvbuf-size 2M;<br/>
unplug-watermark 16;<br/>
}<br/>
}<br/>
</div>
<div>Ressource isn't connectet because need the other Server for another testings.</div>
<div>But finally both would are connected via 10GBe</div>
<div> </div>
<div>Here some outputs from dd:</div>
<div> </div>
<div>Bare metal</div>
<div># dd if=/dev/vgKVM.OS/test of=/dev/null bs=1M count=10240<br/>
10240+0 Datensätze ein<br/>
10240+0 Datensätze aus<br/>
10737418240 bytes (11 GB, 10 GiB) copied, 13,6046 s, 789 MB/s<br/>
</div>
<div>and with DRBD Device:</div>
<div># dd if=/dev/drbd100 of=/dev/null bs=1M count=10240<br/>
10240+0 Datensätze ein<br/>
10240+0 Datensätze aus<br/>
10737418240 bytes (11 GB, 10 GiB) copied, 47,678 s, 225 MB/s<br/>
</div>
<div>Thanks Alex</div>
</div>
</div></div></body></html>