<pre><font size="2">When corroborating data on the actual interface, should the 5 minute input/output rates from a 'sh int' be fairly close to what is reported by MRTG?</font><br><br><br></pre>