As you scale your Observability Pipelines architecture to cover your different use cases:

• If you want to run multiple pipelines on a host so that you can send logs from different sources, follow the instructions in Run Multiple Pipelines on a Host.
• Observability Pipelines uses backpressure signals and buffering to handle situations where the system cannot process events immediately upon receiving them. See Handling Load and Backpressure for more information.
• When you scale Observability Pipelines Workers, each Worker operates independently. See Best Practices for Scaling Pipelines for the recommended aggregator architecture.