<div dir="ltr"><span style="font-family:arial,sans-serif;font-size:13px">Hi...</span><div style="font-family:arial,sans-serif;font-size:13px"><br></div><div style="font-family:arial,sans-serif;font-size:13px">That's another problem running around Cluster Administration on Proxmox...</div>
<div style="font-family:arial,sans-serif;font-size:13px"><br></div><div style="font-family:arial,sans-serif;font-size:13px">I set two VirtualBox VM's running latest PVE version...</div><div style="font-family:arial,sans-serif;font-size:13px">
My laptop, an Intel Core i5 running Ubuntu, act as a Storage with TGT Target...</div><div style="font-family:arial,sans-serif;font-size:13px">I am able to create the cluster and define the quorum disk...</div><div style="font-family:arial,sans-serif;font-size:13px">
However, when I reboot both nodes, I get this error:</div><div style="font-family:arial,sans-serif;font-size:13px"><br></div><div style="font-family:arial,sans-serif;font-size:13px">Starting qdiskd [ FAILED ]... </div><div style="font-family:arial,sans-serif;font-size:13px">
No local IP Address has been set... </div><div style="font-family:arial,sans-serif;font-size:13px"><br></div><div style="font-family:arial,sans-serif;font-size:13px">I think something with DLM lock or something similar issue...</div>
<div style="font-family:arial,sans-serif;font-size:13px"><br></div><div style="font-family:arial,sans-serif;font-size:13px">But, if I go to CLI and check:</div><div style="font-family:arial,sans-serif;font-size:13px"><br>
</div><div style="font-family:arial,sans-serif;font-size:13px"><div>pve01:~# /etc/init.d/cman status</div><div>qdiskd is stopped</div></div><div style="font-family:arial,sans-serif;font-size:13px"><br></div><div style="font-family:arial,sans-serif;font-size:13px">
<div>pve01:~# /etc/init.d/rgmanager status</div><div>rgmanager is stopped</div></div><div style="font-family:arial,sans-serif;font-size:13px"><br></div><div style="font-family:arial,sans-serif;font-size:13px">On both nodes, cman and rgmanager are dead!</div>
<div style="font-family:arial,sans-serif;font-size:13px"><br></div><div style="font-family:arial,sans-serif;font-size:13px">So, if yype the sequence command bellow:<br></div><div style="font-family:arial,sans-serif;font-size:13px">
<br></div><div style="font-family:arial,sans-serif;font-size:13px">/etc/init.d/cman start<br clear="all"><div><div>/etc/init.d/rgmanager start</div><div><br></div></div><div>On both nodes, the cluster go on-line....</div>
<div><br></div><div>I had experience with this issue in physical machines two...</div><div><br></div><div>First, I tought that could be a problem with VirtualBox VM's but it is not...</div><div><br></div><div>So, as a workaround, I put this command in rc.local:</div>
<div><br></div><div>/etc/init.d/cman stop<br clear="all"><div>/etc/init.d/rgmanager stop</div></div><div><br></div><div>/etc/init.d/cman start<br clear="all"><div>/etc/init.d/rgmanager start</div></div><div><br></div><div>
in order to bring cluster on-line...</div><div><br></div><div>Here's the cluster.conf:</div><div><br></div><div><div><?xml version="1.0"?></div><div><cluster config_version="35" name="CLUSTER"></div>
<div> <cman expected_votes="3" keyfile="/var/lib/pve-cluster/corosync.authkey"/></div><div> <quorumd allow_kill="0" interval="3" label="quorum" tko="10" votes="1"></div>
<div> <heuristic interval="3" program="ping <a href="tel:192.168.1.100" value="+551921681100" target="_blank">192.168.1.100</a> -c1 -w1" score="1" tko="4"/></div><div> <heuristic interval="3" program="ip addr | grep eth0 | grep -q UP" score="2" tko="3"/></div>
<div> </quorumd></div><div> <totem token="54000"/></div><div> <clusternodes></div><div> <clusternode name="pve01" nodeid="1" votes="1"></div><div> </clusternode></div>
<div> <clusternode name="pve02" nodeid="2" votes="1"></div><div> </clusternode></div><div> </clusternodes></div><div> <rm></div><div> <failoverdomains></div>
<div> <failoverdomain name="serverfailover" ordered="1" restricted="0"></div><div> <failoverdomainnode name="pve01" priority="1"/></div><div> <failoverdomainnode name="pve02" priority="2"/></div>
<div> </failoverdomain></div><div> </failoverdomains></div><div> <pvevm autostart="1" vmid="100"/></div><div> </rm></div><div></cluster></div></div><div><br></div>
<div>And /etc/default/redhat-cluster-pve has the content:</div><div><br></div><div>FENCE_JOIN="yes"</div><div><br></div><div><div>After running this:</div><div><br></div><div><div>/etc/init.d/cman stop</div><div>
/etc/init.d/cman start</div><div>/etc/init.d/rgmanager stop</div><div>/etc/init.d/rgmanager start</div><div>/etc/init.d/pve-cluster stop</div><div>/etc/init.d/pve-cluster start</div><div>/etc/init.d/pveproxy stop</div><div>
/etc/init.d/pveproxy start</div></div><div><br></div><div>My cluster get on-line, but the more weird issue is here:</div><div><br></div><div><div>clustat </div><div>Cluster Status for CLUSTER @ Thu Jul 10 11:43:30 2014</div>
<div>Member Status: Quorate</div><div><br></div><div> Member Name ID Status</div><div> ------ ---- ---- ------</div>
<div> pve01 1 Online, Local, rgmanager</div><div> pve02 2 Online</div><div> /dev/block/8:33 0 Online, Quorum Disk</div>
<div><br></div><div> Service Name Owner (Last) State </div><div> ------- ---- ----- ------ ----- </div>
<div> pvevm:100 pve01 starting </div></div><div><br></div><div>I remove such VM, 100... It doesn't exist anymore.... But stiil there, according to clustat!!!</div>
<div><br></div><div>Seconds after, I run clustat again and got this message:</div><div><br></div><div><div>pve01:~# clustat </div><div>Cluster Status for CLUSTER @ Thu Jul 10 11:43:46 2014</div><div>Member Status: Quorate</div>
<div><br></div><div> Member Name ID Status</div><div> ------ ---- ---- ------</div><div> pve01 1 Online, Local, rgmanager</div>
<div> pve02 2 Online</div><div> /dev/block/8:33 0 Online, Quorum Disk</div><div><br></div><div> Service Name Owner (Last) State </div>
<div> ------- ---- ----- ------ ----- </div><div> pvevm:100 (none) recoverable </div>
</div><div><br></div><div><br></div><div>And finally:</div><div><br></div><div><div>clustat </div><div>Cluster Status for CLUSTER @ Thu Jul 10 11:44:07 2014</div><div>Member Status: Quorate</div><div><br></div><div> Member Name ID Status</div>
<div> ------ ---- ---- ------</div><div> pve01 1 Online, Local, rgmanager</div><div> pve02 2 Online, rgmanager</div>
<div> /dev/block/8:33 0 Online, Quorum Disk</div><div><br></div><div> Service Name Owner (Last) State </div>
<div> ------- ---- ----- ------ ----- </div><div> pvevm:100 (pve01) failed </div>
</div><div><br></div><div>But, again, there is no VM... </div></div><div><br></div><div><br></div><div>Is there something I do wrong?</div></div><div><br></div>-- <br>Gilberto Ferreira<br>
</div>