If I'm running ingestors, queriers and a compactor across multiple instances, which instance am I supposed to specify the plugin-dir

[jstirnaman]

Enabling processing engine https://docs.influxdata.com/influxdb3/enterprise/plugins/#set-up-the-processing-engine|instructions are not clear on enterprise:

To enable the Processing engine, start your InfluxDB server with the --plugin-dir option
If I'm running ingestors, queriers and a compactor across multiple instances, which instance am I supposed to specify the plugin-dir? All of them? One of them? If one, which one? Does it matter?

Same when setting up https://docs.influxdata.com/influxdb3/enterprise/visualize-data/grafana/#Copyright|Grafana:

URL: Provide your InfluxDB 3 Enterprise URL:
What is my URL? It looks like it can be a querier or ingestor url but not a compactor.

Slack Message