so, no one experienced similar problem, or topic is very rare? Mickael, You know erlang well...any thoughts?<br><br><br><div><span class="gmail_quote">On 9/14/06, <b class="gmail_sendername">Zbyszek Żółkiewski</b> <<a href="mailto:email@example.com" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)">
</a>> wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><div>Helo,<br><br>Second time i experience strange situation with ejabberd running on multiple nodes (3). I have describe that on bugtraq few months ago.
<br>If we have 3 nodes (ejd) that work as cluster and are interconnected, and we break connection to - let say - first node for some period of time, and fix that connection again, nodes will not reconnect. That situation lead to very very unpredictable behaviours - because -
f.e.load balancer "see" that all nodes are working correctly, and it directs connections to that nodes, that lead to strange situations like, wrong s2s negotiations and many others.<br><br>So my question is: is that bug in ejabberd or in erlang itself?
<br><br>If in erlang - is there any way to "force" nodes to periodically check if offline nodes (that look like they are offline - but they are not) come back to life?<br clear="all"><br>-- <br>pozdrawiam,<br></div>
</span></div></blockquote></div><br><br clear="all"><br>-- <br>pozdrawiam,<br>Zbyszek Żółkiewski