<p>We have quite a number of users that run large (>500 agent) call-centers and use the clustering features of QM by maintaining multiple instances of the same queues on clustered, often geographically dispersed systems, using different distribution logic, and using QM to have the reporting as if it was all handled on one single virtual box. </p>
<p>Of course they don't get "true" FIFO call distribution logic, but from what I see they seem to be pretty much satisfied with the results. <br></p><p>It would be cool to have a standard way to have distribuited queues in Asterisk, and if somebody comes up with it, we are surely going to support it in QM. Or one can use the de-facto industry standard and install ViciDial :-)</p>
<p>Thanks</p><p>l.</p><p></p><p><br></p><div class="gmail_quote">2009/1/16 Alex Balashov <span dir="ltr"><<a href="mailto:abalashov@evaristesys.com">abalashov@evaristesys.com</a>></span><br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
QueueMetrics is a great product and I highly recommend it.<br>
<br>
The main problem with QueueMetrics from my point of view has nothing to<br>
do with QueueMetrics whatsoever; on the contrary, it solves the problem<br>
it intended to solve very well.<br>
<br>
It's just that at the point at which the kind of analysis and reporting<br>
it provides becomes most useful, a single hostwise queue instance is<br>
likely to be inadequate. Queues need to exist on multiple hosts. And<br>
since Asterisk doesn't have distributed queues at this time, one has to<br>
adopt other means of getting around that problem. And those means don't<br>
tend to be compatible with the existing queue infrastructure or allow<br>
one to easily use QueueMetrics.<br>
<div class="Ih2E3d"></div></blockquote></div><br>-- <br>Loway - home of QueueMetrics - <a href="http://queuemetrics.com">http://queuemetrics.com</a><br><br>