Thanks in advance for any suggestions/instincts/intuitions.
First, I am not an Asterisk developer - so really I can only put this in general terms.
I just would like to consult the fine gentlemen (and any ladies) on this forum for any possible ideas or suggestions as to the [possible] causality of the effects being observed - as things just are not making sense.

Okay here we go:

Say an asterisk server has been set up at call center SITE(A).
Everything has been configured, tested and running fine for, oh, a week.
Have both pumped calls in simulated as well as live tested SITE(A) up to 50 human agents.
The server and card is new and robust - there is no lacking in spec.
Purring like a kitten.
Now we have another call center at SITE(B) with 50 agents (in relativley close geographical proximity)
What happens is this:
Proceed to cut-over SITE(B) call center and agents to server at SITE(A)
Observation at SITE(B) at first is fine.
Within a short while (say a few minutes) - there enters considerable LAG - 3000ms or more
Additionally inbound-call cannot register and the entire list becomes UNREACHABLE
Then - intermittely, the lag will improve to normal/acceptable time eg. 6ms-15ms
This is at SITE(B)
Here is the intersting part
Somehow SITE(A) - having been tested and purring like a kitten for more than a week exhibits the same symptoms and problems - call-ins cannot register and REACHABLE (and stable for a week) becomes UNREACHABLE.
Now, theoretically - SITE(B) location really shouldn’t matter - this is TCP/IP and server is just a dumb box withouth bias or prejudice to packets from countries with different religious or politic.
Simulation cannot be conducted from an alternate site as SIP trunk lines are physical and unique to SITE(B).
It’s self-evident this is either a network or server issue.
High and intermittent lag times suggest it is a network issue.
Checking with the NOC - they say no problems and the weather is beautiful - and some more - if there were problems their phones would be ringing off the wall - which they are not.
Which points to the server being the problem. Load? Configuration? Nope. Already tested and working just fine for more than a week on SITE(A).
Some developers will be doing their stuff, checking TCP logs etc.
I am just posting to this forum, again, with interest to hear what the many Asterisk Pros out there - given a very general and non-technical explanation - may intuit as the potential issue that can result is such (odd) behavior.
Of course, I know there are so many things that any developer would need to know to be able to pinpoint the problem - again - just consulting this forum to hear very general and potential ideas of what ghost-inside-wired is having fun toying around.
Many thanks to All.
And Happy Holidays!

We’re going to need a lot more specific stuff. If you say a phone can’t register, show us the SIP logs so we can see why it failed. Since you haven’t said otherwise, I’ll assume you are using the signalling built right into Asterisk (instead of, say, openSIPS).

Thanks, redwolf890, for your reply and sharing your instinct that, given the vague and incredibly undetailied explanation, points towards the engine and closely inspecting the SIP logs; will see what I can learn by asking the developers about the setup later today [IF] I get with them (no promises). Additionally, PING utility and similar to server at SITE(A) work fine from SITE(B) - PING times are of 2-5ms on average. Anyways, in case anyone is wondering, I work more on business development not the technical side, although I have a little background in IT - and was on site yesterday and observed this happening - so I’m just taking the initiative and efforting to help out “on-the-side” with whatever I can. Sometimes developers are so into their work they would not think to post to a forum, similarly, as we all have heard and possibly experienced ourselves, the male species, forutnately, unfortunatley, is not always inclined to pull over at a gas station and ask for directions. :laughing: