right now, Conduit-Manager provides per-container bandwidth limits (e.g., via settings) that apply to individual Conduit containers.
However, there is no option to define a global total bandwidth limit across all running Conduit instances.
This would allow node operators to cap the aggregate network throughput for all Conduits running on a server — which is helpful to:
Prevent network saturation on servers with limited uplink capacity
Ensure fair usage when serving many peers
Protect other services sharing the same network connection