vicidial.org

I now have the signature of my setup!

Anyway, in my continuing efforts to improve our setup we have ran into a bit of a snag. We changed to a SIP provider from PRI lines and all of our agents have noticed an increase in "dead air" calls. According to the data the dead air calls are terminated by an agent over three times more often than a caller. Its definitely a telling number. This is something agents complained about when we converted our new system to SIP (the cluster setup). We just recently changed over our other office and they started complaining about the dead air as well. Two separate instances not knowing about each other complaining about the same thing after the switch.

So my first question would be, does anyone have some ideas on what I can analyze for this problem?
My second question would be. Is there a way that I can increase the rate at which the script refreshes for the agents? We have a very powerful database server and I was thinking if we could just query it more often maybe I could shave milliseconds off the time between when the agent gets a call and notices that they get a call (as mentioned in previous posts, I removed the "bloop" sound when people enter vicidial conferences, so an agent only knows they have a call when the script changes).

If you need any more information just ask. Thanks in advance!

Have you tried switching to a different SIP provider? Different carriers have different terminating capacities. This might minimize the "dead air" you are getting. You might also be getting calls that has already ended.

No, we just started using SIP. I think it might be in our best interest to test other SIP providers.

Increasing the refresh has not been necessary for anyone to date, but if you wanted to try it you could certainly ask The Vicidial Group to double the refresh rate in a screen for you (or create a refresh rate preference so you could play with it). Consider creating a ticket in the "Issue Tracker" (Link at the top of the screen) and seriously consider sponsoring it. If your mysql server, agent workstations, and network can handle the load.

Advice given to all clients in our colo and in general: THREE carriers minimum. Paychecks ride on this, there is no excuse for "oops, the carrier is down so you can't feed your kids today, sorry". LOL

Seriously: All carrier's go down eventually. Maybe only for 20 minutes ... but when you're IN that 20 minutes you do not know how long it will be. So ... do you send everyone home? After 10 minutes? 30? An hour? Hm. Change an 8 to a 9 and *poof* another carrier goes live and quite often the agents never even know it happened.

Thanks william. I think the issues Im having is a whole different beast. But I will get additional SIP providers, working on that now.

Of course it's also entirely possible that the missing bloop is causing agents to be "blissfully unaware" and then the client hangs up because there's no sound ...