OpenVidu WebRTC platform allows developers to build awesome videocall applications in minutes.
Madrid, Spain
Joined October 2016
- Tweets354
- Following36
- Followers745
- Likes52
Pinned Tweet
OpenVidu 3.9.0 is out!
It brings a renovated OpenVidu Meet with programmatic control over live meetings, and a better OpenVidu Platform that behaves more predictably under network changes, high load and node restarts. Dozens of bug fixes complete it.
openvidu.medium.com/openvidu…
We interviewed Janne Sikiö, CEO at Dicode, about why "Just use Teams" was not enough for them.
Dicode's non-profit clients needed secure video calls that truly protected sensitive health data.
So they built a video solution with privacy at its core. 🛡️
What's next for Dicode's platform? 🔮
They're looking at call recording and AI summaries, but following their privacy-first model:
AI models must run locally, never exposing confidential data to external parties.
OpenVidu makes this level of privacy AI possible. 🤖
Ready to build private, scalable, and fully customizable video calls? 🚀
Dicode's success story shows how you can build solutions that truly respect data privacy.
Discover how they did it, and learn to empower your own applications:
👇
openvidu.io/blog/2026/10/27/…
🤯 Imagine testing a production-scale distributed system on a single laptop!
That's the challenge when building complex platforms like OpenVidu.
We needed a way to replicate entire clusters and even break them on purpose. 💥
This "cruelty" has uncovered obscure, critical bugs:
🐢 Slow Redis taking down healthy nodes
🔄 Configuration changes silently reverting
👻 Recordings stuck as "active forever"
It means a more robust OpenVidu for everyone. 💪
Want to see how we build robust, fault-tolerant video solutions on a laptop? 🧑💻
Learn how Docker-in-Docker enables true chaos testing.
Discover the techniques for your own distributed systems!
👇
Read the full article: openvidu.io/blog/2026/10/20/…
Recordings failing... almost 100%!" 🚨 Customer's WebRTC cluster had errors on nearly every recording. The twist: 😮 Metrics showed low CPU, ample headroom. No demand spike, no overload. What was going on? Something deeply wrong. 🤯
Key takeaways from this "black hole" bug: 💡
🔑 Complex systems have surprising second-order effects.
🔑 Metrics don't always tell the full story; logs are crucial.
🔑 Robust observability is vital to trace hidden failures.
This kind of story teaches us the most! 📚
This bug story highlights the importance of deep observability and smart system design.
It's fixed in OpenVidu 3.6.0!
👇
Ready to dive into the full details, see the diagrams, and learn how we traced this bug step-by-step?
Read the full article: openvidu.io/blog/2026/10/13/…
