<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">

<html>

<head>

</head>

<body bgcolor="#ffffff" text="#000000">

Hi all,<br>

<br>

I'm having an issue with a setup using the following:<br>

cluster-glue-1.0.6-1.6.el5.x86_64.rpm<br>

cluster-glue-libs-1.0.6-1.6.el5.x86_64.rpm<br>

corosync-1.2.7-1.1.el5.x86_64.rpm<br>

corosynclib-1.2.7-1.1.el5.x86_64.rpm<br>

drbd83-8.3.2-6.el5_3.x86_64.rpm<br>

kmod-drbd83-8.3.2-6.el5_3.x86_64.rpm<br>

openais-1.1.3-1.6.el5.x86_64.rpm<br>

openaislib-1.1.3-1.6.el5.x86_64.rpm<br>

pacemaker-1.0.9.1-1.el5.x86_64.rpm<br>

pacemaker-libs-1.0.9.1-1.el5.x86_64.rpm<br>

resource-agents-1.0.3-2.el5.x86_64.rpm<br>

<br>

This is a two-node HA cluster, with the nodes interconnected via bonded

interfaces through the switch. The issue is that I have no control of

the switch itself, can't do anything about that, and from what I

understand the environment doesn't allow enabling multicast on the

switch. In this situation, how can I have the setup functional (with

redundant rings, rrp_mode: active) without using multicast.<br>

<br>

I've seen that individual network sockets are formed between nodes,

unicast sockets, as well as the multicast sockets. I'm interested in

knowing how will the lack of multicast affect the redundant rings,

connectivity, failover, etc.<br>

<br>

I've also seen this page<br>

<meta http-equiv="content-type" content="text/html; charset=utf-8">

<a

 href="https://lists.linux-foundation.org/pipermail/openais/2010-October/015271.html">https://lists.linux-foundation.org/pipermail/openais/2010-October/015271.html</a><br>

And here it states using UDPU transport mode avoids using multicast or

broadcast, but it's a patch, is this integrated in any of the newer

versions of corosync?<br>

<br>

Thank you in advance.<br>

<br>

Regards,<br>

<br>

Dan<br>

<pre class="moz-signature" cols="72">-- 

Dan FRINCU

Systems Engineer

CCNA, RHCE

Streamwide Romania

</pre>

</body>

</html>