<br><br><div class="gmail_quote">On Thu, Jan 27, 2011 at 3:24 PM, Felix Frank <span dir="ltr">&lt;<a href="mailto:ff@mpexnet.de">ff@mpexnet.de</a>&gt;</span> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<div><div></div><div class="h5">On 01/27/2011 11:13 AM, Nauman Yousuf wrote:<br>
&gt; here<br>
&gt;<br>
&gt; Primary DRBD<br>
&gt;<br>
&gt; skip {<br>
&gt;         Lustre storage cluster replication setup<br>
&gt;         LUSTRE OSS active/active Replication<br>
&gt;<br>
&gt; }<br>
&gt;<br>
&gt; global {<br>
&gt;     minor-count 2;<br>
&gt; }<br>
&gt;<br>
&gt; resource r0 {<br>
&gt;   protocol C;<br>
&gt;<br>
&gt;   startup {<br>
&gt;     degr-wfc-timeout 120;<br>
&gt;   }<br>
&gt;<br>
&gt;   disk {<br>
&gt;     on-io-error   detach;<br>
&gt;   }<br>
&gt;<br>
&gt;   net {<br>
&gt;   }<br>
&gt;<br>
&gt;   syncer {<br>
&gt;     rate 30M;<br>
&gt;     group 1;<br>
&gt;     al-extents 257;<br>
&gt;   }<br>
&gt;<br>
&gt;   on <a href="http://typhoon.example.com" target="_blank">typhoon.example.com</a> {<br>
&gt;     device     /dev/drbd0;<br>
&gt;     disk        /dev/cciss/c0d0p8;<br>
&gt;     address    <a href="http://10.65.200.21:7788" target="_blank">10.65.200.21:7788</a>;<br>
&gt;     meta-disk  /dev/cciss/c0d0p6[0];<br>
&gt;   }<br>
&gt;<br>
&gt;   on <a href="http://cyclops.example.com" target="_blank">cyclops.example.com</a>{<br>
&gt;     device    /dev/drbd0;<br>
&gt;     disk      /dev/cciss/c0d0p8;<br>
&gt;     address   <a href="http://10.65.200.30:7788" target="_blank">10.65.200.30:7788</a>;<br>
&gt;     meta-disk /dev/cciss/c0d0p6[0];<br>
&gt;   }<br>
&gt; }<br>
<br>
</div></div>I&#39;m not familiar with Lustre operation. Does /dev/cciss/c0d0p8 mean that<br>
both nodes in fact use the same chunk of Lustre data? Because that would<br>
probably be bad.<br></blockquote><div><br>NO each machine have separate c0d0p8 which is part of drbd then drbd device are part of lustre  .. to have a combined SAN. that i user for lustre client . <br></div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">

<div><div></div><div class="h5"><br>
&gt; resource r1 {<br>
&gt;   protocol C;<br>
&gt;<br>
&gt;   startup {<br>
&gt;     degr-wfc-timeout 120;<br>
&gt;   }<br>
&gt;<br>
&gt;   disk {<br>
&gt;     on-io-error   detach;<br>
&gt;   }<br>
&gt;<br>
&gt;   net {<br>
&gt;   }<br>
&gt;<br>
&gt;   syncer {<br>
&gt;     rate 30M;<br>
&gt;     group 1;<br>
&gt;     al-extents 257;<br>
&gt;   }<br>
&gt;<br>
&gt;   on <a href="http://typhoon.example.com" target="_blank">typhoon.example.com</a> {<br>
&gt;     device     /dev/drbd1;<br>
&gt;     disk       /dev/cciss/c0d0p9;<br>
&gt;     address    <a href="http://10.65.200.21:7789" target="_blank">10.65.200.21:7789</a>;<br>
&gt;     meta-disk  /dev/cciss/c0d0p7[0];<br>
&gt;   }<br>
&gt;<br>
&gt;   on <a href="http://cyclops.example.com" target="_blank">cyclops.example.com</a> {<br>
&gt;     device   /dev/drbd1;<br>
&gt;     disk      /dev/cciss/c0d0p9;<br>
&gt;     address   <a href="http://10.65.200.30:7789" target="_blank">10.65.200.30:7789</a>;<br>
&gt;     meta-disk /dev/cciss/c0d0p7[0];<br>
&gt;   }<br>
&gt; }<br>
&gt;<br>
&gt;<br>
&gt; Secondary DRBD<br>
&gt;<br>
&gt; skip {<br>
&gt;         Lustre storage cluster replication setup<br>
&gt;         LUSTRE OSS active/active Replication<br>
&gt;<br>
&gt; }<br>
&gt;<br>
&gt; global {<br>
&gt;     minor-count 2;<br>
&gt; }<br>
&gt;<br>
&gt; resource r0 {<br>
&gt;   protocol C;<br>
&gt;<br>
&gt;   startup {<br>
&gt;     degr-wfc-timeout 120;<br>
&gt;   }<br>
&gt;<br>
&gt;   disk {<br>
&gt;     on-io-error   detach;<br>
&gt;   }<br>
&gt;<br>
&gt;   net {<br>
&gt;   }<br>
&gt;<br>
&gt;   syncer {<br>
&gt;     rate 30M;<br>
&gt;     group 1;<br>
&gt;     al-extents 257;<br>
&gt;   }<br>
&gt;<br>
&gt;   on <a href="http://typhoon.example.com" target="_blank">typhoon.example.com</a> {<br>
&gt;     device     /dev/drbd0;<br>
&gt;     disk        /dev/cciss/c0d0p8;<br>
&gt;     address    <a href="http://10.65.200.21:7788" target="_blank">10.65.200.21:7788</a>;<br>
&gt;     meta-disk  /dev/cciss/c0d0p6[0];<br>
&gt;   }<br>
&gt;<br>
&gt;   on <a href="http://cyclops.example.com" target="_blank">cyclops.example.com</a> {<br>
&gt;     device    /dev/drbd0;<br>
&gt;     disk      /dev/cciss/c0d0p8;<br>
&gt;     address   <a href="http://10.65.200.30:7788" target="_blank">10.65.200.30:7788</a>;<br>
&gt;     meta-disk /dev/cciss/c0d0p6[0];<br>
&gt;   }<br>
&gt; }<br>
&gt;<br>
&gt; resource r1 {<br>
&gt;   protocol C;<br>
&gt;<br>
&gt;   startup {<br>
&gt;     degr-wfc-timeout 120;<br>
&gt;   }<br>
&gt;<br>
&gt;   disk {<br>
&gt;     on-io-error   detach;<br>
&gt;   }<br>
&gt;<br>
&gt;   net {<br>
&gt;   }<br>
&gt;<br>
&gt;   syncer {<br>
&gt;     rate 30M;<br>
&gt;     group 1;<br>
&gt;     al-extents 257;<br>
&gt;   }<br>
&gt;<br>
&gt;   on <a href="http://typhoon.example.com" target="_blank">typhoon.example.com</a> {<br>
&gt;     device     /dev/drbd1;<br>
&gt;     disk       /dev/cciss/c0d0p9;<br>
</div></div>&gt;     address    <a href="http://10.65.200.21:7789" target="_blank">10.65.200.21:7789</a> &lt;<a href="http://10.65.200.21:7789" target="_blank">http://10.65.200.21:7789</a>&gt;;<br>
<div class="im">&gt;     meta-disk  /dev/cciss/c0d0p7[0];<br>
&gt;   }<br>
&gt;<br>
&gt;   on <a href="http://cyclops.example.com" target="_blank">cyclops.example.com</a> {<br>
&gt;     device   /dev/drbd1;<br>
&gt;     disk      /dev/cciss/c0d0p9;<br>
&gt;     address   <a href="http://10.65.200.30:7789" target="_blank">10.65.200.30:7789</a>;<br>
&gt;     meta-disk /dev/cciss/c0d0p7[0];<br>
&gt;   }<br>
&gt; }<br>
&gt;<br>
&gt; I am using DRBD to replicate Lsuter data .<br>
<br>
</div>How? Where is your MDS? What are your OSSs? Where are the nodes that use<br>
DRBD and how do they access Lustre? (Sorry, I cannot infer that from the<br>
/dev/cciss/... stuff).<br>
<br></blockquote><div> /dev/ciss is particular partition on typhon and cyclops which make a DRBD in between to have a raid device on netwoek then that raid is part of lustre to have redundant disks.<br></div><div> </div>
<blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Also, please be careful to CC the list.<br>
<br>
Regards,<br>
<font color="#888888">Felix<br>
</font></blockquote></div><br><br clear="all"><br>-- <br>Regards<br><br>Nauman Yousuf<br>0321-2549206<br>E-Eager, N-Noble, G-Genuine, I-Intelligent, N-Natural, E-Enthusiastic, E-Energetic, R-Resourcefull --- ENGINEER<br>