Follow-up Plan for Resource Management Improvements

[seetadev]

Once IdentifyPush and Yamux are completed and reviewed, we can start planning the resource management work. Here's a rough outline of the steps:

• Design Phase

Review how go-libp2p's rcmgr works and identify which concepts are most critical for py-libp2p.

Define a minimal viable resource manager (MVRM) for py-libp2p — something lightweight to start with, but extensible.

Identify key resource types to track: streams, connections, memory usage, inbound peer counts, etc.

• Implementation Phase

Introduce basic enforcement of configurable limits (connections per peer, streams per connection, max memory use).

Build a simple internal metrics system to track resource usage (or integrate with an external one like Prometheus-compatible exporters).

Add hooks for detecting and reacting to abusive peers (e.g., automatic blocking, stream resets).

• Testing and Hardening

Simulate common attacks: stream spamming, payload flooding, sybil connections.

Tune default limits for safety without being too restrictive.

Test behavior under both normal and adversarial network conditions.

• Future Enhancements

Implement autoscaling limits based on system load (CPU/memory).

Integrate resource stats into a monitoring dashboard.

Add support for dynamic peer scoring and connection gating.