I'm pretty new to K8 and cannot find any information for my kind of problem.

I have an algorithm that is running in a pod in K8 on an Edge device (Linux worker node). The Edge device runs Windows Server 2019. Hyper-V is used for the Linux worker node's VM.

When the pod containing the algorithm container is started, not all data, which the algorithm will process, is available. Data will be created by another machine (medical device) in the network and should to be streamed into the container.

Mounting a volume into the pod seems to be not appropriate for this because of the performance loss. We do not want to start the pod only after all data is available. We want to start the pod right after the first data is available and then continuously send the new data to the pod.

Is there a method to send data streams or chunks of data to a pod?

Please let me know in case my description is not clear. Thanks for your kind support!