Vodia PBX

Well, that timeout happens after the call is already over. The gateway initiates the hangup, and it says it does that because of RTP problems. You can see in the BYE response that the PBX sends far less packets than it receives. So it is understandable why the gateway hangs up. Maybe you have silence suppression turned on? With media related problems, it makes sense to get a Wireshark trace. Then we will be able to see why the PBX does not send RTP data. Maybe it is because of the Speech Engine, maybe there is a option to disable silence suppression. Also, check if there is a upgrade availale. 3.0.0 sounds like there is...

Could be a problem with the ACK routing (do you have more than one IP address or a firewall somewhere?). Or could be a tone detection that is too aggressive detecting a hangup. Maybe you can get the SIP packets between the AC and the PBX and we can see if the hangup comes from there. If that does not give any insight, we can do the same thing between the PBX and Exchange. Divide and conquer.

I don't have the Mitel phone here, but you should set up username, password, domain, outbound proxy as usual. If you plan to install a large number of handsets, we can look into PnP.

Is there any reason why you are using 8.3? I remember some of the 8 versions were not very good. What is the latest? 8.6? Do you have a chance to give that a shot? From the logs I must say it seems that the phone has some problem authenticating. I does not respond to challenges and that is really strange.

Well, considering that a single DID is a dollar value every month I see the registration problem relaxed. The issue is that ther eis not clear specification on how to deal with one registration and multiple identities behind it. Especially because the CS410 will probably have only a few DID per box I would say lets take it easy in the beginning and then later when the standards are more clear a trunk may have more than one DID. Actually I know carriers who seriously provide hundreds of DID to the same client each of them with a seperate registration.

I would factory-reset the phone to make sure that there is no residual configuration on the phone. Then use the BLF mode on the buttons. That should work just fine.

I remember testing Mitel phones some time ago (well, must be more than a year...). But anyway, it worked and I don't remember any significant problem. The only disadvantage I can think of is the PnP will not work out of the box.

Once you have an audio session enabled you can re-INVITE with all kinds of video and image codecs. The PBX will just be the SBC between the endpoints then, and H.263 can be transmitted between the connected phones.

What kind of support are you referring to (http://wiki.pbxnsip.com/index.php/Trouble_Ticket_Processing)? I assume this is 3.0.0.2998. A agent group does not have a voicemail. If you want that the call goes to voicemail, you have to specify a timeout value and use as destination the voicemail of an extension. Usually, you can put a "8" in front of the extension number. For example: After hearing ringback for (s) ...: 30 ... redirect the call to the destination (e.g. "73"): 8123 That means that after 30 seconds of ringing an agent the call will be redirected to the mailbox of extension 123. This setting has nothing to do with a ringing agent. This timeout is for the waiting time in the queue (where a caller would hear music on hold). 5 seconds if extremly short, a typical value would five minutes (which is 5 * 60 = 300 seconds). Call forward on no answer is a setting that affects calls directly to the extension. If the call comes in through a agent group, that setting has no meaning (otherwise if every agent sents their own little call redirection the chaos would be complete). Adding features has always the danger of making things complicated, escpecially with a relatively complex thing as a waiting queue. On the one side we get the pressure to add features here and there, and on the other side we get complaints that the software is difficult to use. We try to take the advice from both sides and have more features, but at the same time still make it possible for an average person to use them. It is not always easy.

Retrieve is the same thing.

Maybe you are right. Another was of generating kilobytes in order to turn a light on (after dialog).

Okay, but then you don't have to worry about caller-ID on the CS410. You can use ANI, or a trunk prefix or just s DID for the whole trunk. Just like any other installation. On the hosting side, you would have one "extension" for each DID number. In NAPNA, I would say one DID one registration, that keeps things very simple.

In the bad.rtf there is a call redirection turned on on the phone to 18606704381?!

I don't want to be too optimistic here. Presence has the problem that you might see the "presence" (whatever that is), but the PBX is not able to tell the phone what to do when the button is being pushed. And there are a lot of x-cisco-xxx headers in the packets. Look for "extended-refer", if you want to have some fun. When the draft was written Rohan Mahy was still with Cisco, that is a long time ago. I would not build opon that...

BLF is supposed to be really stupid. It does show the status of the monitored resource. But whenever you are pushing the button, it just calls the number. Pickup is not possible with that, that is a feature because it avoids race conditions (incoming call and at the same time pushing the button will accidentially pick up an incoming call - you don't want that). Speed dial is even more stupid. It just dials the number, no matter what. Transfer scenarios in SIP require that the phone initiates a REFER. That is a little bit tricky with the buttons. Lets see if a firmware upgrade on the phones addresses this problem in the future.

I dont 100 % get it... You mean the CS410 is only for the case of failover? Or should it run at all times?

I think the easiest way to solve that problem is to use the setting "Explicitly specify park orbit preference". If someone put a "*" in there it will mean "ask". This way we nicely stay backward compatible with that we have now and we can even specific the behavior on per-extension basis.

If you can, give 3.0.1.3016 a try: http://www.pbxnsip.com/protect/pbxctrl-3.0.1.3016.exe.

Check out http://wiki.pbxnsip.com/index.php/Snom_M3.

There is no settings for that, but you can easily fool it by putting a prefix to the cell phone and then strip that prefix in the dial plan. For example, if your cell phone number is 212-123-4567 you would put 999-212-123-4567 there and then have a dial plan entry with the pattern 999*.

Well for park/retrieve, that is not so easy. There is already a park/retrieve code that does not take an extension behind it. Plus in this case there is a useful meaning behind it - let the PBX search for you. And usually there are not so many park orbits, so that it is reasonable to put one on each key.

Okay, the next step is to get packet level trace (e.g., Wireshark) to see what is going on on the cable. Then we see if it is a jitter problem, packet loss problem, SRTP problem or whatever.

Next version will have a command line option "--no-check-ports" that will allow the PBX to start up even if FATAL errors are reported. TFTP and SNMP will not be fatal any more. The default is that the PBX will not start up any more if the SIP or HTTP ports cannot be used by the PBX.

Well, that is a policy question... If the PBX does not start because it does nto get the TFTP port, I would call that picky. HTTP is also not essential for making phone calls... So where is the line? IMHO it is reasolable to start the process anyway. If you have HTTP, then you can fix the other problems through the web interface. If you have SIP, you can run the service already.

Well, those errors are probably because the PBX does not have superturtle powers to open those protected ports.