<div dir="ltr">In the same moment I stopped keepalived the memory stabilize.<div><br></div><div>I attach the tcpdump output for 61613 port. ( it should be just keepalived connections from the other balanced</div><div>server, as this cluster has no queues )</div>
</div><div class="gmail_extra"><br><br><div class="gmail_quote">On Fri, Sep 27, 2013 at 3:49 PM, Emile Joubert <span dir="ltr"><<a href="mailto:emile@rabbitmq.com" target="_blank">emile@rabbitmq.com</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br>
Hi.<br>
<div class="im"><br>
On 26/09/13 15:18, Santi Nuņez-Cacho wrote:<br>
<br>
> =INFO REPORT==== 26-Sep-2013::16:12:46 ===<br>
> accepting STOMP connection <0.19254.0> (XX:49217 -> YY:61613)<br>
><br>
> =INFO REPORT==== 26-Sep-2013::16:12:46 ===<br>
> closing STOMP connection <0.19254.0> (XX:49217 -> YY:61613)<br>
<br>
</div>I've attempted to reproduce this problem from your description but did<br>
not succeed. Can supply a network packet capture file of TCP traffic<br>
between the broker and the load-balancer under the conditions that lead<br>
to the leak? Tools such as tcpdump can create such a file.<br>
<span class="HOEnZb"><font color="#888888"><br>
<br>
<br>
<br>
-Emile<br>
<br>
<br>
<br>
<br>
<br>
</font></span></blockquote></div><br></div>