We Love Streaming Video on Kubernetes and We are About to Tell why

Flussonic is a standalone solution that cancels video delivery issues from top to bottom. That’s why we never get tired of saying that Flussonic is an all-in-one tool for working with video streaming.

For proving it month after month, with each new version of Flussonic that we deliver, we have been included in the top 50 companies in the industry according to Streaming Media.

Each new version of Flussonic Media Server is created to make the user experience even more enjoyable and administration even easier.

Not only developer, but also DevOps friendly

For the last few years we have set ourselves the goal of making our software “developer friendly” and we are convinced that we are on the right track. Updating the software on a regular basis allows us to constantly bring new features to our customers and our ability to create something customized to suit any particular need is also proof of that.

This year we also made a complete transition to OpenApi, so even the most complex integrations can be done in record time. Now the instructions are clearer, the API methods are perfectly described and linked to the “schema”.

Also this 2022, more precisely this Fall, we broke into the world of DevOps by adapting Flussonic for Kubernetes. Thanks to this it is now possible to manage a group of servers without personal participation, update them with a button, scale dynamically and set parameters for which you need to add nodes. Our users will finally be able to sleep peacefully at night as everything is available in a single solution.

It is now possible to launch Flussonic in the Kubernetes environment and have a server up and running ready to receive posts in no time.

Another product we are developing is Flussonic Cloud, a streaming video service. Kubernetes is used for the cloud (we have real operational experience), and this is very exciting for us. We think you will like it too!

Contact us today, we will tell you how our Cloud is managed through Kubernetes and we will help you with your service.

How was the adaptation?

Below we will tell you about some of the adaptations made, which were simply necessary to ensure full compatibility with k8s.

We had to figure out what to do with the activation files. We found a way for Flussonic to put them in a local database: Kubernetes secret storage.

The key, login and password “arrive” from the outside, from the cluster setup. That is, without folding in git, without manually entering the password. Those who don’t need to have the cluster’s production passwords won’t know them. Keeping files secret is also good because it allows purchased Flussonics to be independent of license servers for a while: everything will work 24 hours a day.

One of the necessary adaptations was also the ability to log records in JSON format. Kubernetes immediately saves the logs to its storage, where you can safely view them: “I know what to do with the logs, just give them to me.” Using standardized tools like this will make implementation cheaper. First of all, you don’t need to waste time figuring out where to write these logs, install separate software on a separate log collection server, which would read them, download them periodically; then other software that would unpack, put in a database, compress, index…

Second, if you write logs to a disk and don’t deliver them, and you have 50 Flussonics running, you’d still have to guess where to look for the logs for the desired stream instead of seeing everything in one central place.

Third, resources on the hard drive would be consumed. Each of these 50 Flussonics would record their own logs, so it’s also hard to predict space usage.

Anyway, you can forget about this nightmare and not think about it anymore. Logs are now written “centrally” and conveniently.

Guides

The first thing to remember is that we always have and still have support for running in a container. As well as instructions on how to do it.

Regarding the configuration in Kubernetes, we have already put a guide in the documentation. In the public git repository you will find sample yaml files and you can create a cluster for your publication. There is also an example of autoscaling, so that you serve your peaks, and outside of them – there was no overspending of resources.

We will not stop here. In the near future, for example, a separate guide will appear specifically for UGC services, making it even more convenient.

–

P.S. It is worth mentioning separately that all Flussonics modules work on k8s. Including our famous DVR with the ability to add depth to the archive for over a year and a seamless transition from viewing recording to live streaming. Or, for example, a transcoder with the ability to prepare a multi-bitrate stream for stable playback for viewers with slow Internet access.

Use it and run it with peace of mind. Everything works. We have checked!