Logicwrath

I just meant from the community searchbox from this website. --- As far as my solution for dynamic outbound caller id.  I can post it here for now. Create a call flow using Postman and the API.  Here is an example of the call flow data: http://pastebin.com/js3uba9q Edit the following php file with the caller id's you want to match. http://pastebin.com/hAsdSQqU You can set a logging variable to dump the headers and the other important variables to a text file. The call flow dialing pattern should match anything with a * and 10 or 11 digits.  *15558675309 So instead of pressing 1 for country code people can do * and use the dynamic matching pivot script.  I have not put this into production yet.  I stopped looking at it when it worked a few times.

Thank you for your review/attention!

It's worth repeating, that the phone still shows DNS-NAPTR and not DNS-SRV.

I manually entered the 0, it was port 7000 originally.

Changing the outbound proxy port to 0 caused the phone to failover to us-east correctly!

https://dl.dropboxusercontent.com/u/23868996/syslog.tar

Let me know when you have it so I can take it down.

It still says DNS-NAPTR.  Here is the screenshot of the firmware and hardware revision.  http://prntscr.com/ah46xb Here is the relevent account configuration. http://prntscr.com/ah47dr Here is the capture.  https://dl.dropboxusercontent.com/u/23868996/28.72.0.45_5_12_22.pcap

Do you want the us-central to be blocked or unblocked during this capture? I am guessing blocked.  Kind of obvious! :)

I also just tried changing my DNS settings on the phone from Windows Server DNS to OpenDNS in case it was related.  Same thing.

I tried 28.73.0.45 (on accident) and then 28.72.0.45.  In both cases I let the phone register and then stop block us-central in the forward chain of my router.  The T46G will not fail over.

I was also thinking if firmware versions become relevent for proper continuity should we consider letting the provisioner app automatically upgrade or downgrade firmware of the device to ensure it is in a supported state?  I believe this should be possible pretty easily and you would ensure that everyone using a specific model phone is in a supported firmware.  This is especially interesting when you think about drop shipping phones, you will really have no idea what firmware will come from the distributor. You could always have an option to disable for people requiring specific firmware.l

Sure.  That was with firmware 28.80.0.60.  I just upgraded to 28.80.0.95 and tried again.  I track all traffic with destination port 7000 in real time in the mikrotik.  Same thing. The reason i said 10-15 seconds previously is because as soon as I block the IP I will try placing a call to see how quick it recovers.  If I have the second outbound proxy configured It will drop the call in 3-6 seconds the next call I place will typically go through.

This is what I get, it is selecting NAPTR.

Here is my experience.  I let the device get the defaults from Advanced Provisioner at both the account settings and the device level.  Reset T46G to default, autoprovision, and get a clean configuration.  Provision from http://p3.zswitch.net. *.s.zswitch.net us-central.p.zswitch.net I then go into my mikrotik firewall create an address list blocking the us-central IP on the forward chain and completely disable the phone.  It will timeout, and then display invalid account because it is not getting the other servers back from DNS.  If I manually setup a second outbound proxy to us-east the phone and run the same test with the firewall the phone will almost immediately disconnect stop trying on us-central and failover to us-east.  I can also place a call within 10-15 seconds. So how do I make the default configuration work without adding a second outbound proxy?  I assumed the NAPTR and SRV combo setup would force the phone to get the priority servers based on the main address you configured (us-central) however, using the default zswitch DNS and realm and advanced provisioner I am unable to fail over this T46G automatically when I block its connection to the single us-central proxy.

This sounds great.

Yes, setting up 20-30 phones and manually entering 4 fields for each line key/BLF by hand when they are 99% the same is super inefficient.

Yes, I can test this out no problem.  Also, I found an issue where the provisioner is not setting the second outbound proxy on my Yealink T46G's.  It will say that the correct setting is going to be inherited for the second server, but it will not show up.  It will also not configure it if you manually enter the second outbound proxy, versus letting it inherit.   Not sure if that was on your bug list for this next release.  It requires us to manually update the outbound proxy in each phone, which negates using the provisioner for saving time.  I just tested this again last night.

http://prntscr.com/agz5oa It appears this might be a T46G only setting, it has a large screen and we have to set that to extended to see entire line key name if it is longer than like 4-5 characters as the default length will cut it off.

Also, I did fix the branding issue.  Under the field "Domain Name" I was using my domain name, not my portal domain (CNAME) name.  This was not clear.  This is an example of something I wish I could fix, whether it be via editing community articles or contributing back to the code base.  I would have been happy to edit the branding article in the community to make this more clear.  There are other examples of documents that could be improved.  This is why I was hoping to get a user wiki or something we have access to improve. I appreciate the helpful documents, FAQs, etc.. I just wish we could somehow improve them as a community.  I suspect you guys are too busy to be building out articles and documents all the time.

This is still happening.  I walked my account rep through this issue earlier in the week.  Open Accounts Manager settings in the app store, change access to specific user, choose yourself, logout and log back in, you will not be able to access the account manager app. Seems like a bug.  If you can't reproduce, I will create a video but my account rep reproduced it earlier this week over the phone.

Yes, this was sorted out via the account rep.  It is all set.  Thanks!

That is great news on the Fax ATA settings.  I was hoping someone would post their best practices on those.  It seems you are already working on it.  Awesome. I have been very happy with the T37 email faxing features built into Smart PBX. I think the HTTPS based ATA's are really where the community needs to be going in terms of using an on-premise based fax machine.

I think your first point is great.  The more people contributing the better.  It would also be nice if people could earn these credits for other things besides specifically contributing code.  I want to impact the community, however, I may not be qualified to write code for Kazoo.  At the same time I may be able to provide helpful posts, documentation, and/or code snippets (think pivot) that overall improve the community.  Perhaps these non-code related contributions might be worth some support credit or bug fixes in the future if you think that makes sense.  In any case I think you are on the right track.

Yes, this works for me.  It likely will not be searchable via the community, however, that may not be necessary for people to use it.