mc_

You may want to investigate the phone's REGISTER sequence here. We've seen some softphones that send a REGISTER for 300s then immediately follows up with a REGISTER with no Expires header which KAZOO treats as a deregister.

Just to be an agent of chaos, work commitments have changed such that Wednesday 10 isn't a viable time now (and actually Tuesday at 9:30 has opened back up again). So Tuesday 9:30??? :)

Leaning Wednesday 10:00 pacific myself, least often to have a conflicting meeting these days. We'll see if more folks trickle in with requests; will update the calendar here on the forum once we solidify something.

The module will look for a child branch in "children" that matches the value found at "variable"'s path in the "scope" data structure. So if "scope" is "custom_channel_vars" and "variable" is "authorizing_type", the module will look for children based on "authorizing_type" (like "device", "resource", etc) in the call's "custom_channel_vars" JSON object

Some ideas here: https://docs.2600hz.com/dev/applications/callflow/doc/cf_branch_variable/

Firewalls with SIP ALG typically target UDP (and maybe TCP) packets destined for port 5060 on the remote side. When ALG can't be turned off at the client site, moving traffic to port 7000 (a typical alternative SIP port) makes the packets pass through the firewall unmolested. Adding TLS obviously encrypts the packets to keep them from being mangled as well - not required to move to 7000 to do this. Not sure if there's a standard or convention, but I think TLS-enabled connectivity targets ports 5061 and 7001 on KAZOO: https://github.com/2600hz/kazoo-configs-kamailio/blob/69840f0c5a11237b1b216072e411d3d98df006eb/kamailio/listener-defs.cfg#L9

What is KAZOO doing at this time? There are routines that run monthly for updating account balances and things like that; could also be that your server is underpowered

Check /etc/kazoo/bigcouch/local.ini the database_dir config: https://github.com/2600hz/kazoo-configs-bigcouch/blob/master/bigcouch/local.ini#L16

The API is the callflows API. These are actions within the "flow" of a callflow.: https://docs.2600hz.com/dev/applications/crossbar/doc/callflows/ About callflow structure: https://docs.2600hz.com/dev/applications/callflow/doc/

Your best bet for a system reset is, on each couch server, delete everything in /srv/db then restart kazoo-bigcouch. Then restart kazoo-applications to re-initialize system databases.

The virtual KAZOOCon this year should have some announcements in that regard.

Have you compared the call logs of the working and non-working calls? I'm surprised the a-leg's CAVs have your value in the answered call, and don't have the CAV in the unanswered. I would expect the CAV to be set on the a-leg even before the PIVOT request is made. Of course, click2call might be messing that up since it does use loopbacks.

Any in particular? A few are exported across the bridge but most pertain to the leg itself and shouldn't be exported. You can see the bridge strings have the exported CCVs on them when creating b-legs.

Not sure how I missed the PR! I will comment over on it for a couple things and see if we can get it included.

use_local_resources:true will check the account for carrier docs; use only if the account itself has carriers defined. DIDs on the system are forced outbound by either carrier module = knm_local OR "force_outbound" set to true (I think just those two). Try `sup stepswitch process_number {DID}` and see if the output includes forcing the DID offnet. If using local resources, I think you can add the account ID as an argument after {DID}.

When the endpoint with "push" settings is being built, the code doesn't know yet if the SIP device itself is registered or reachable yet. So the "push" options are delayed 5s to allow the SIP device a chance to be contacted. If there's no registration or the network doesn't work to deliver the call to the SIP device, the "push" device is called (meaning push notification sent and Kamailio is ready and waiting for the SIP device to register its new location on the network.

When you have push configs on a device, KAZOO will create two endpoint configs for FreeSWITCH/Kamailio - the "real" device and the "push" device. KAZOO gives the "real" device 5s to answer the call before initiating the push notification to allow for network shenanigans.

These are MODBs: https://docs.2600hz.com/dev/core/kazoo_modb/doc/ They are critical to KAZOO usage, so there's no disabling of them. If the MODB doesn't exist when a document (like a CDR) needs to be saved, the MODB will be created. If you have a lot of activity, they'll all stack up waiting for the MODB to be created. What you can do is configure the MODB task (https://docs.2600hz.com/dev/applications/tasks/doc/modb_creation/) Set that to pre-create the next month's MODB for all accounts and give it enough lead time (say 3-5 days depending on how many accounts you have); the task will automatically spread out the MODB creations across that time period to hopefully minimize impact on the DB cluster.

Don't think so, offhand. There's a 'set_audio_level' FS command that could be useful? If you want to try it out and let us know, we can look into incorporating it

As far as I recall, KAZOO has some facility for coverting wav<->mp3 but that is rudimentary and I think only for uploading media through the API. Your best bet is to fetch the recording and convert it on your server.

You'll likely need to tinker with FreeSWITCH settings; I'm not sure if alternatives are supported though. Probably post-recording format conversion will be needed.

A basic intro to the API and setting up devices and carriers: https://docs.2600hz.com/sysadmin/doc/install/configure_kazoo/ From there, check out the bigger collection of Voice APIs: https://docs.2600hz.com/dev/ KAZOO does not have LCR capabilities, just basic ratedeck support. Most folks setup LCR servers as a carrier for KAZOO to route through when needing that functionality.

KAZOO does not do outbound registrations, meaning it will not register to your peerless account. One way would be to create a "device" in KAZOO to represent the peerless system and have peerless register to KAZOO (if possible). That would allow extension dialing. You can also create peerless as a carrier in KAZOO and accept inbound calls from it (peerless IP(s) need to be in the ACLs); but you would need to setup E164 numbers on your callflow. So a little more complicated but would set you up to accept traffic from other carriers as well.

You'll need to investigate the db side (maybe check HAProxy logs too). Is couch on a public IP? Maybe someone is messing with you? Check kazoo logs, make sure you aren't inadvertently deleting the account through some automation gone wrong.