<div dir="ltr"><br><div class="gmail_extra"><br><br><div class="gmail_quote">2014-08-12 13:19 GMT+02:00 Bastian Scholz <span dir="ltr"><<a href="mailto:nimrodxx@gmx.de" target="_blank">nimrodxx@gmx.de</a>></span>:<br>
<blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">Sounds like some kind of network problem...<br>
<br>
I assume that your three switches are normal<br>
managementable switches?<br></blockquote><div><br></div><div><br></div><div>HP pro curve 2510G-24 (switch A)<br></div><div>zyxel GS1510-24 (switch B)<br>zyxel GS1910-24 (switch C)<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>
Maybe you have (an old) Spanning Tree Protocol<br>
enabled<br></blockquote><div><br></div><div>Switch B is the only one with STP enabled.<br> <br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
But when the whole Network went bad under load<br>
it could be a bad uplink cable, bad port or a<br>
strange switch failure... Can you reproduce the<br>
bad behavior with other traffic (netcat for<br>
example)?<br></blockquote><br></div><div class="gmail_quote">I can't because this is a production environment.<br></div><div class="gmail_quote">Now I also mode the last interface of vmserver006 under switch A.<br></div>
<div class="gmail_quote">I disconnected the firewall and no node got disconneted.<br></div></div></div>