Hi,
Are there any recommended memory and cpu lower limits when running the Meeting SDK container in Kubernetes pod to join a zoom meeting as a bot? Our plan is to fetch the audio bytes as then come in and hand them over to another server, so that we don’t use too much memory in the pod where the container is running.
@chunsiong.zoom
Thanks
@pkenia without video, I will still go for the minimum specs of 2 core 4gb ram.
Good thing on Kubernetes is that you need to only set the upper limit. It might make sense to do a test to see what’s the peak consumption. The biggest portion cpu and ram consumption would be probably from processing of raw audio. You might want to experiment with that to see how you can fine tune that.
I’m thinking saving pcm to wav file might be a efficient way, since bandwidth and storage is relatively lower cost compared to processing
Hey @pkenia, based on my experience, the suggested specs seem reasonable, and they might even be a bit over-provisioned depending on your hardware and specific use case. From what we’ve observed at Recall.ai with our Zoom native bots, we rarely exceed 1 core of CPU and 1.2 GB of memory, and that’s while processing video—which tends to be the most resource-intensive part. Since you’re only handling audio, you should need even fewer resources than we do. Another option is the using Recall.ai API. It’s a simple 3rd party API that lets you use meeting bots to get raw audio/video/metadata from meetings without you needing to spend months to build, scale, and maintain these bots. Here are the API docs: https://docs.recall.ai
Thanks @chunsiong.zoom and @amanda-recallai .