<br><div class="gmail_quote">On Fri, Sep 17, 2010 at 11:22 AM, Nyamul Hassan <span dir="ltr"><<a href="mailto:mnhassan@usa.net">mnhassan@usa.net</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
While this is too many "eggs" in one basket, but can be useful if you have "too many" E(T)1s say equivalent to a STM1 (OC3) or more. In that case, it would be too many boxes at 8ports / box.<div><br></div>
<div>Somewhere in the mailing list, Sangoma devs said that they do 32E(T)1 per box on the labs quite frequently, although mostly for load testing.</div><div><br></div><div></div></blockquote><div><br></div><div>That is correct, that is our typical load test scenario. However, Asterisk is a complex system with many features. Our testing focus on SIP to TDM bridging, meaning the only used applications are Answer() and Dial() with the DAHDI and SIP channel drivers, typically with latest 1.4.</div>
<div><br></div><div>Additionally we always compile DAHDI modifying the chunk size to reduce the interrupt load.</div><div><br></div><div>As far as your question about PCIe 2.0, yes the A108 should work just fine there.</div>
<div><div><br class="Apple-interchange-newline">Moises Silva<br>Senior Software Engineer<br>Sangoma Technologies Inc. | NEW 100 Renfrew Drive, Suite 100, Markham ON L3R 9R6 Canada<br>t. 1 905 474 1990 x128 | e. <a href="mailto:moy@sangoma.com" target="_blank">moy@sangoma.com</a></div>
<br><br></div></div>