Repository navigation
Release v8.0-d1
Pre-release
If you would like to support SRS, please consider contributing to our OpenCollective.
c7b89ff56cc139096f7a7207197afa5d28155f3f
Release v8.0-d1, 8.0 dev1, v8.0.63, 356563 lines.
Add the v8.0-d1 release line to the README releases and the SRS 8.0 changelog.
Co-Authored-By: Claude Opus 5.5 noreply@anthropic.com
Resource
- Source: 8967aa549c794bb5d9e68a4fee883c20 srs-server-8.0-d1.tar.gz
- Binary: 978833e0220ce47405b3c569546b17ea SRS-CentOS7-x86_64-8.0-d1.zip
Docker
Usage
Performance: v8.0-d1 vs v8.0-d0
Each scenario is an A/B run: both builds at once in Docker, each SRS on its own vCPU, the same load and the same flat CBR source (H.264 200 kbps + AAC 32 kbps, 25 fps). SRS CPU is averaged over 180 s and compared second by second. Negative means v8.0-d1 uses less CPU. Changes within about ±1 point are noise.
| Scenario | Clients | v8.0-d0 CPU | v8.0-d1 CPU | Change (95% CI) |
|---|---|---|---|---|
| RTMP → RTMP | 11,520 | 50.2% | 48.5% | −3.4% (−4.3% to −2.5%) |
| RTMP → HTTP-FLV | 11,520 | 75.0% | 70.3% | −6.2% (−6.5% to −5.9%) |
| RTMP → HTTP-TS | 2,880 | 54.1% | 53.7% | −0.7% (−1.4% to −0.1%) |
| RTMP → RTSP | 2,880 | 67.4% | 64.8% | −4.0% (−4.2% to −3.7%) |
| RTMP → HLS | 5,760 | 13.3% | 13.5% | +0.8% (−2.4% to +4.1%) |
| RTMP → DASH | 5,760 | no media¹ | 16.9% | — |
| RTMP → WHEP | 1,152 | 59.4% | 60.7% | +2.3% (+2.1% to +2.6%) |
| WHIP → WHEP | 1,800 | 74.0% | 75.5% | +2.1% (+1.8% to +2.4%) |
| WHIP → RTMP | 11,520 | 54.2% | 49.6% | −8.5% (−10.6% to −6.5%)² |
| WHIP → HTTP-FLV | 11,520 | 77.2% | 74.5% | −3.6% (−4.1% to −3.1%) |
| SRT → SRT | 576 | 41.5% | 41.7% | +0.5% (+0.3% to +0.8%) |
| SRT → RTMP | 5,760 | 19.2% | 22.4% | +16.7% (+15.0% to +18.3%) |
| SRT → WHEP | 1,152 | 66.3% | 67.7% | +2.2% (+1.7% to +2.7%)² |
| RTMP publishers | 1,152 | 83.0% | 82.1% | −1.2% (−1.5% to −0.9%) |
| WHIP publishers | 400 | 19.0% | 19.6% | +2.9% (+2.5% to +3.2%) |
| Many streams (1 pub + 2 players each) | 576 streams | 66.9% | 65.5% | −2.1% (−2.3% to −2.0%) |
¹ v8.0-d0 players received no DASH media (every player failed); v8.0-d1 served all 5,760 players.
² Both builds briefly went over 2 s of player lag in this run, so the load was near its limit; the comparison is still like for like.
- Faster: HTTP-FLV (−6.2%), WHIP → RTMP (−8.5%), RTSP (−4.0%) and WHIP → HTTP-FLV (−3.6%).
- Slower: SRT → RTMP +16.7%.
- Fixed: DASH playback now works under load.
- No real change: everything else is within about ±3%.
Setup: Mac mini M4 (Mac16,10), Docker 29.8.2, Linux 7.0.14 VM with 8 GB. These numbers compare the two versions on one machine; they are not the capacity of a production server.