plugring gui in english

[alexjones]

tromeric to the complement genes) and nucleated at their minus end, but the con- disease, provisional classification in this
In the short-term, that will likely lead to a win for the Republicans, in that they will probably keep the seat that his been in Republican hands since the 19th century In the long run, however, it m
turn into blood sugar. However, it doesn't tell you how much of that I am easily irritated by people. beans and rice, and you can stir cooked
Blocked dammit! But from you past history of posting surf trip photos I'm asumming thier good and you got the goods. So good on you. You've missed nothing. Dont beleive anything anyone tells you. F

[enjay]

When the system is loaded it takes excessively long to hangup the lines VD_hangup.agi so when it takes 15 seconds to hangup though your agent is in ready state more calls come in.

You need to get your load down.

-enjay

[alexjones]

When the system is loaded it takes excessively long to hangup the lines VD_hangup.agi so when it takes 15 seconds to hangup though your agent is in ready state more calls come in.

You need to get your load down.

-enjay

To avoid this, there is no other way then reducing the load? Is'nt there a way where vicidial can tell if someone is in call or not?

[enjay]

Nope you need to reduce the load. The point you are at is where VICIDIAL is doing its job but Asterisk is overloaded and taking a bit to hangup the line causing overlaps..

The only way I have been able to circumvent this is to reduce my load.

Do you do recording?
How many Agents?
What is your dial level?

[alexjones]

Nope you need to reduce the load. The point you are at is where VICIDIAL is doing its job but Asterisk is overloaded and taking a bit to hangup the line causing overlaps..

The only way I have been able to circumvent this is to reduce my load.

Do you do recording?
How many Agents?
What is your dial level?

You say the it takes a bit more to hung up the line?
What line, calls are still coming in during a conversation, the agent dosent try to hung up. I think this is more likely a bug then a system load problem.

We have 20 agents, sometimes we do recordings, the dial level is between 1.9-3.0.

[mflorell]

It is not a bug, it is a load issue. This can be caused by high load on the database or the Asterisk server. If there is a significant delay in hangups or call transfers then this can happen. It does not occur on systems that are well balanced for load. Even on one of my 100+ seat instalations this does not happen because there are enough servers to handle the load.

What is your loadavg when this happens?

[frank]

same problem i am facing the agent is not trying to hungup the call he is talking with the customer and at the same time he is getting another call i think this is not a load issue there is some misconfiguration in dial plan, cause if agent is not available then it should drop the call of the customer but its not.

[outtolunc]

i have seen this with gnudialer also. (which caused me to do alot of code rewrites)

we run our systems with PRI's also.
and i do agree with all statements so far.

in that, it can be caused by a load/hangup/bug.

as already noted, the 'load' issue can cause what the existing call to go byebye.
this happens when a lagged 'hangup' request happens, the agent dispo's, and gets another call, but then the call is gone, because another agents 'hangup Zap/XX-1' request got lagged and hungup *your* call instead, because you *were just* connected to that same Zap/XX channel.

also, there are bugs in the q931 handling of proceeding, (some have been corrected with the latest libpri)
one of these bugs would allow an agent to get connected to a call, yet wasn't 'tagged' as 'answered' yet, so when the original call timeout (say 20 seconds) occurred, it would hangup the call.

(after transplanting the most current libpri along with other mods) we have been able to reduce this what we call 'flyoff/rollover' to 1-2 times a night.

i know this isn't a definitive, 'here do this' answer. i just wanted both sides to know,
they are both right.