I would appreciate hearing about your experiences with using Talk with the high-performance backend. How many video streams can you run simultaneously without any lag? Where are the bottlenecks that you might not think of?
Background to my question: we have a NGO with fairly complex network layout with VPN/Intranet/public network access, Wi-Fi, various locations, a heterogeneous end device landscape and a powerful bare metal server on which the HPB runs via Sunweaver script. Monitoring of the server shows no abnormalities in the logs and very low utilisation. The HPB connection also seems to be running smoothly in Nextcloud. In practical tests, video conferences start to stutter at around 12-15 streams.
Your practical experience may help us to find and eliminate the bottlenecks. How many streams run smoothly in parallel for you?
The Basics
- Nextcloud Server version (e.g., 29.x.x):
- Nextcloud Hub 9 (30.0.16)
- Operating system and version (e.g., Ubuntu 24.04):
- Nextcloud-Server:
Ubuntu 24.04 on Kubernetes
- Nextcloud-Server:
- HPB on Debian latest installed only with Sunweaver-Script (and Metrics), Server with 12 cores, 64GB RAM, 1GBit/s network - BareMetal in public network
- Are you using CloudfIare, mod_security, or similar? (Yes / No)
No
I would appreciate any helpful information. Thanks a lot
Jo