Jump to content

mc_

2600Hz Employees
  • Posts

    1,767
  • Joined

  • Days Won

    4

Posts posted by mc_

  1. If you create a user with any fields not defined in the schema, KAZOO will store them untouched. So you can easily add "team_id" to the user doc when you create it and you will get it back when you fetch the user doc.

    For the CDR portion, you should get the Owner-ID included if the device involved in the call leg is owned by a user. Assuming you track the KAZOO User's ID + "team_id", you should be able to receive the CDR, read the Owner-ID and associate the CDR with the user / team in your application.

    And welcome! :)

  2. You can use the conferences API to dial out to the participant - this will create the conference if not already running and put the callee into it in one step.

    But you can't create dynamic conferences with a statically defined callflow - you need Pivot to help you return the conference callflow JSON with the conference name you want to use.

    Perhaps if you tell us more about your logic for how you name the conferences, maybe there's a way to do it?

  3. Nope, there's no "target" (which is basically a callflow extension/number) named conference1 so it won't work. You could transfer to "target":"pick-conference" then create a callflow with "numbers":["pick-conference"] and a "flow":{"module":"pivot",...} that will call out to your server. You do your logic, realize the caller needs to be connected to "conference1" so you return {"module":"conference", "data":{"config":{"name":"conference1"}}}

    Something along those lines should get you where you want to go.

  4. Obviously the web server, on port 3333, is responding. However, your initial report was the UI telling you it couldn't connect to the server, which is the KAZOO API server. That server address should be stored in the config.js file somewhere (forget the format but should be obvious).

    The KAZOO URL is the one not reachable. Test that...then check kazoo logs, make sure KAZOO and Crossbar are running on the expected server(s), etc.

  5. @fmateo05 thanks, its a team effort with 2600Hz, our partners, and the community at large. Thank you for participating here on the forums, I know lots of people are benefiting from your questions, comments and answers!

    Oh yeah, if SF in March sounds feasible, anyone can get a discounted ticket by using my code "JamesAimonetti15". I get no kickbacks here, you just get 15% off! :)

  6. @Ken Rowland it is my expectation that using the 'set_variables' callflow action with '"data":{"custom_application_vars":{"my":"secret"}}' will cause all subsequent call events for this call leg to include '{"my":"secret"}'. If you add '"export":"true"` then the bridged call leg should also carry them.

    If that's not the case, please open a support ticket so we can investigate. But we do have customers that rely on these CAVs being present and, AFAIK, we have not received additional reports of missing CAVs.

  7. July 28th 2010, the first commit to an ambitious project, Whistle, was made.

    We're coming up on 10 years of KAZOO and all its history! Holy moly!

    Part of the fun will be a talk I'm giving in March in San Francisco at the CodeBEAM conference. I'll be going through some of the history of KAZOO, how we got where we are today, some war stories, horror stories, hopefully some levity and minimal amounts of 20/20 vision jokes. But come on, there's probably going to be at least one! As a dad, I can't not make 'em!

    I would love for folks to come, especially anyone interested in (or already doing) Erlang as this is a premier conference to attend. The hallway track is worth the price of admission and the talks are typically top-notch (excepting yours truly but apparently I know where the bodies are hid?).

    Whether you can join the conference or not, I'd also love to hear your stories around this project and maybe include some in the talk if you permit. Are there things you want to know more about, like technical decisions, that I can make sure to speak to in my talk? Anything that sticks out as "I'm really glad I'm part of this community"?

    Look forward to hearing from you all!

  8. The first implementation of Pivot was a basic TwiML clone, so we probably didn't check the content-type response header then. When we added KAZOO's callflow JSON as an option, Pivot started checking the content-type to know which backend, kzt_twiml or kzt_kazoo, to use interpreting the response.

    Now, why your server changes from text/xml to text/html I can't answer. Its obvious the preceding responses had their content-type header set to text/xml so you would need to check the code path taken on your side for this failing response to see why it didn't set the content-type header.

  9. Pivot posted to ' Jan  7 16:05:57 pintransfer 2600hz[1137]: |call-711CE674-8013-3810-000C-2F@72.82.244.65|pivot_call:431(<0.2507.0>) sending req to http://209.142.178.205:57113/Keefe_2//Keefe/amountEntered(post): Digits=1&CallerNumber=9940140000&CallerName=9940140000&Direction=inbound&ApiVersion=2010-04-01&ToRealm=165.227.107.0&To=9940140000&FromRealm=165.227.107.0&From=9940140000&AccountSid=72bd25b60a91cfebe5bf1f50894126ee&CallSid=call-711CE674-8013-3810-000C-2F%4072.82.244.65 '

    That returned text/html which is wrong, it should be text/xml, as the previous responses returned.

  10. @bsimpson72515  the previous Pivot requests to your URLs all returned with 'Content-Type: text/xml'. 'text/xml' is associated with the twiml backend of Pivot and so life is good. The errant response itself is fine, XML-wise, but the server is returning 'Content-Type: text/html' which is not associated with any Pivot backend. Why does the web server response with the proper Content-Type header in all but the last response?

×
×
  • Create New...