Hello community,
I am currently evaluating using nextcloud talk for a non profit organisation that has more than 500 volunteers, in order to have meetings where there are:
- 1 person doing screen sharing at a time
- 2-5 active mics at most
- 1-10 web camera feeds
- Rest of the 500 people will be passive and will consume the content
While calculating a bit on the requirements, I understand that this will take a significant amount of bandwidth, sending out the stream to all the consumers - I think I got up to 1.2 - 1.5 gbps through the high performance backend, as well as quite a few CPU cores and RAM.
Since I am piloting this and don’t have much experience with nextcloud talk at this scale, I am willing to do my best to see if it could work out.
My questions are:
- Are there anyone else here that has experience with the high performance backend and the other components at this scale? I would love to hear what your challenges has been and how you ended up solving them. After searching, I did not find any meaningful information.
- How well does the backend scale with multiple instances at once? Are there any modes to have it load balance between multiple instances of the backend, to share the load?
These meetings do happen every quarter (approximately) and so my initial plan was to rent a very powerful cloud node with significant bandwidth to take the load for the few hours it is needed.
I feel like i am stumbling into a usecase that is extremely uncommon for most users, so I am also happy to share my learnings along the way in this thread, for future queries like these.