Are logs from separate files combined into the same pipeline? re: "reduce" transform

[davidcpell]

Hopefully this question makes sense. I want to use "reduce" to combine Go panic logs in our vector agents, but wondering if I can assume whether logs from separate files are kept separate within the vector process with their own copies of the component pipeline. Does each file tailed in kubernetes_logs get its own pipeline in memory or are all the logs from the various files mixed together going through the same instances of each component?

[pront]

Does each file tailed in kubernetes_logs get its own pipeline in memory or are all the logs from the various files mixed together going through the same instances of each component?

That depends on your pipeline but generally there's no implicit pipeline creation. If you define kubernetes_logs source A then all events produced by this source will go to the same stream. So if you later have a reduce transform T that has A as input, it will receive all those events.