<div dir="ltr"><span class="gmail-"></span><br><span class="gmail-"></span><div class="gmail_extra"><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><span class="gmail-">
<br>
<br>
</span><span class="gmail-">> for the<br>
> moment. I have done a drbdmanage primary drbdctrl but still the drbd<br>
> storage is not available. How can I resolve the split brain manually so<br>
> as the drbd storage continues to work even if pve1 (primary is down).<br>
<br>
</span>I guess that this should be left to drbdmanaged daemon, i guess the<br>
control node should not be managed directly. May be someone of drbd<br>
folks could give more details/advices.<br></blockquote><div><br></div><div>Ok we shall see if someone from drbd could give some advices here. <br></div><div><br> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<span class="gmail-"><br>
> I will try to test the scenario by adding a third drbd node (pve3) to<br>
> the cluster (drbdmanage add-node command) on pve1 and I will let you know.<br>
<br>
</span>With three node you should avoid the split brain. By the way, even with<br>
a two node pve cluster you would be a degraded situation,<br>
<br>
no quorum ... -> most of the operations locked for safety.<br>
<br>
The same holds for the drbd storage cluster.<br></blockquote><div><br></div><div>Yes thats true. <br></div><div><br> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<br>
By the way, my three node cluster performs as expected, the only<br>
difference between mine and yours being that I have three nodes for drbd<br>
storage as well.<br></blockquote><div><br><br></div><div>I have added the third node (pve3) to the drbd9 cluster but with option --no-storage since I don't have enough space. I am using it only as a control node for drbd9. All the vms will be stored on pve1 and pve2. <br></div><div>I test the HA all seams to be working without any problem. <br><br></div><div>As per my understanding, as from drbd9, the two-nodes drbd server setup is no longer possible when it comes to HA since at the backend its drbd who manages the roles of nodes and allocate the leader. If 2 nodes are up then its ok but when 1 is down among the 2 drbd storage is no longer accessible with drbdmanage. <br></div><div>Could someone please confirm this?<br></div><div><br></div><div>Thanks<br clear="all"></div></div><br>-- <br><div class="gmail_signature"><div dir="ltr"><div><div dir="ltr">Shafeek SUMSER<br></div></div></div></div>
</div></div>