Red5 Pro Release 7.0

Introducing Red5 Pro version 7.0!

The latest release improves audio performance by supporting higher audio bitrates (96 kHz stereo streams at 128 kbps) as well as stereo support (browser dependent). This means you can get better than CD quality audio files over WebRTC.

We also expanded the functionality of the autoscaling solution by adding the ability for a single Terraform server to support multiple stream managers behind a load balancer, a node relations map, and the ability to request a transcoder by region. Furthermore, we streamlined the autoscaler performance with improved connection logic and bug fixes.

Don’t miss our updated server capacity benchmarks and NEW publishing benchmarks.

We would like to thank our hardworking team for always driving to improve Red5 Pro and continuing to keep it a progressive and well-functioning software tool. Read below for the full list of improvements, fixes, and updated benchmarks.

Release 7.0.0, 8 June, 2020

Server Updates

Stereo audio published streams in WebRTC (browser dependent)
Firefox can publish and play back stereo
Chrome publishes stereo but plays back in mono
Safari sometimes publishes stereo
MS Edge does not publish stereo
Support for high audio bitrates.
Ability to support stereo audio over WebRTC. (browser dependent)
Configuration for high quality audio and high bitrates for WebRTC subscribers transcoded from rtmp/rtsp published streams.
Fixed – 6-9 seconds to display video Firefox RTC Publisher to Chrome RTC subscriber
Ice4j Role Conflict Patch (Firefox edge-case scenario)
Fixed – In a cluster, stream sometimes does not show/play on one edge but is fine on another
Fixed – Added a special download to address Red5 Pro not running on Windows Server versions

Autoscaling API 4.0 Updates

Autoscaling API has been updated to 4.0. Please see stream manager upgrade documentation for adding client backwards compatibility
New – Standalone terraform server to support multiple stream managers behind a load balancer see updated documentation
New – Option to target a specific image per node type in launch configuration
New – Node Relations Map, displays relations between nodes in a group in a more graphical manner
New – able to request transcoder by region
Updated SM relay logic so that edges only connect to one relay, to better distribute load
Updated WebSocket proxy implementation for scalability and more stable connections in high load situations
Updated SM response logic for getting multiple subscriber requests at the same time to distribute traffic better amongst multiple edges
Updated SM logic to check for a stream before directing a subscriber to an edge
Updated SM logic to wait for edges to be inservice before directing subscribers
Added support for additional AWS regions, including China and African regions
Fixed – In an autoscaling group with multiple transcoders – streams published to one transcoder are also displaying on other transcoders
Fixed – Republishing same stream name doesn’t return new entry in event/usage api call
Updated stream manager examples added to the streaming HTML5 testbed

Release 7.0.0 Server Performance Metrics

Tests were run against an AWS c5.large instance (2 CPUs with 4GB memory, 2GB allocated to java_heap). We used our RTC, RTSP and RTMP Bee clients to do load testing. Note – connections are added over the course of several minutes.

SUBSCRIBER TESTS

Publishing a 240p (426×240, 200 kbps) stream via RTMP, we were able to achieve the following while still maintaining quality of stream:

500 WebRTC subscribers, or
1,800 RTSP subscribers, or
1,200 RTMP subscribers

Publishing a 480p (854×480, 500kbps) stream via RTMP, we were able to achieve the following while still maintaining quality of stream:

350 WebRTC subscribers, or
1,300 RTSP subscribers, or
900 RTMP subscribers

Publishing a 720p (1280×720 1,500kbps) stream via RTMP, we were able to achieve the following while still maintaining quality of stream:

250 WebRTC subscribers, or
1000 RTSP subscribers, or
700 RTMP subscribers

Publishing a 1080p (1920×1080 4,500kbps) stream via RTMP, we were able to achieve the following while still maintaining quality of stream:

100 WebRTC subscribers, or
700 RTSP subscribers, or
400 RTMP subscribers

PUBLISHER TESTS

The same server type (2 CPUs with 4GB memory, 2GB allocated to java_heap) can support approximately (tested using the RTCBee Publisher test, RTMP Bee Publisher test and RTSP Bee Publisher test):

240p: