AstriDevCon 2020 Agenda
Date¶
Discussion items¶
Time | Item | Who | Notes |
---|---|---|---|
10:15-11 | State of Asterisk Talk | Matthew Fredrickson | Link to presentation: AstriDevCon Asterisk Update - 2020.pdf |
11-11:45 | Asterisk 18 - Codec Work | George Joseph | Link to presentation: AstriDevCon 2020 Advanced Codec Negotiation.pdf |
11:45-1:15 | Lunch | ||
1:15 | Start building detailed agenda |
Afternoon topics:
- Outbound ARI discussion (websocket initiates from Asterisk rather from external agent)
- Are there ways to improve bug reports on the public issue tracker?
- PCAPs
- Core dumps
- sipp scenario files
- Maybe adding a bounty flag on the tracker so that people could figure out which ones have bounties attached.
- What happened to the bridge created event in ARI (not showing up until first channel is added)?
- Handling on -dev list
- Is there additional information that needs to be added to certain debug/error/warning messages to better understand the original source of error?
- Consistent identifying attributes on log messages (IAX, chan_sip) - owner information, callId/reference.
- Add log message per channel driver linking channel_id and channel_name and protocol specific callId/callref
- Would helper functions help enforce this?
- Group photo
- Discussion around challenges supporting queue strategy changes for dynamic environments
- Adapting codec quality to the network (potentially using RTCP feedback or native means)
- Feedback to sender to alter sending
- Handled in codec module (callback exists to provide RTCP information so codec module can adjust)
- End to end sequence number preservation exists now to help determine gaps for codec implementations
- Improved RTCP stats logging
- Ability to disable RTCP messages in AMI
- Would be nice to get RTCP stats in stasisend event (due to not having access in ARI after stasis end)
- Ability to add to CDR log or CEL event log would be neat. Perhaps using custom CDR log.
- Better documentation and easier to find
- Do access to things work in hangup handlers?
- Are there challenges that people have with provisioning fleets of Asterisk instances in the cloud? (missing provisioning APIs, logging interfaces, ...)?
- Nice to have a solid ARI version of app_voicemail that works well across multiple instances (all playing together well with recordings, metadata, etc)
- Challenges with the way files are stored with app_voicemail across multiple instances of Asterisk and shared file stores.
- It would be nice to record files and send them to another server directly from Asterisk (maybe using remote FTP/HTTP storage or something of that nature). We can already playback remotely, why not recording as well?
- What would be interesting to see next?
- Blind transfer across Asterisk instances
- Media failover (between Asterisk instances when one fails or for draining of calls from one misbehaving instance to another).
- PauseRecording/UnPauseRecording support in mixmonitor is missing - recommended workaround is to use mixmonitor stop and then mixmonitor append.
- Better handling of RTP header extensions in Asterisk. (passthrough or processing of radio related header, potentially also for webrtc and other header extensions)
- Optimistic encryption support when using DTLS with chan_pjsip