feat(release): v0.9.3 - Orchestration Intelligence

[imzzaidd]

v0.9.3 -- Orchestration Intelligence

Summary

Introduce intelligent orchestration with PID-controller-based auto-scaling, BFT health consensus, Reed-Solomon erasure coding for storage redundancy, and KZG polynomial commitments for data availability. This release makes the octopus orchestration layer self-tuning and resilient, while upgrading storage and commitment schemes to state-of-the-art approaches.

Planned Changes

PID Controller Auto-Scaling

• Proportional-Integral-Derivative controller for octopus subnode scaling
• Target metric: network load, presence throughput, validator latency
• Anti-windup protection and derivative filtering
• Configurable gains (Kp, Ki, Kd) via governance parameters
• Bounded output clamping (respect INV63 subnode limits)

BFT Health Consensus

• Byzantine fault tolerant health assessment across subnodes
• 2/3+1 agreement required to declare a node healthy/unhealthy
• Health state machine: Healthy, Degraded, Unhealthy, Recovering
• Integration with validator slashing for unresponsive nodes
• Protocol-level view change for leader rotation on health failure

Reed-Solomon Erasure Coding

• RS(n, k) encoding for epoch-bound storage data
• Configurable redundancy factor (default: 2x, meaning k-of-2k recovery)
• Systematic encoding for efficient normal-case reads
• Integration with pallet-storage for automatic encoding/decoding
• Bounded shard size to respect on-chain storage limits

KZG Polynomial Commitments

• Kate-Zaverucha-Goldberg commitments for data availability proofs
• Trusted setup ceremony support (powers of tau)
• Batch opening verification for efficient DA sampling
• Integration with pallet-zk as a commitment scheme option
• Domain separation for KZG evaluations (7ay:kzg:v1)

Deliverables

• PID controller module with configurable gains
• BFT health consensus protocol for octopus subnodes
• Reed-Solomon encoder/decoder (GF(2^8) with Cauchy matrix)
• KZG commitment scheme with batch verification
• Octopus pallet integration with PID auto-scaling
• Storage pallet integration with RS erasure coding
• ZK pallet integration with KZG commitments
• Trusted setup ceremony tooling for KZG
• Performance benchmarks: scaling response time, RS throughput
• BFT health simulation test suite

Invariants Affected

• INV38-42: Octopus orchestration upgraded with PID scaling and BFT health
• INV63: Subnode limit enforcement integrated with PID output clamping
• INV41: State consistency strengthened via BFT health consensus
• INV70-72: Storage integrity enhanced with Reed-Solomon redundancy
• INV73-75: ZK proof system extended with KZG commitment option
• INV46-49: Validator economics updated for health-based slashing

Breaking Changes

• Octopus subnode management API extended with health states
• Storage encoding format changes (raw to RS-encoded shards)
• New KZG commitment type in pallet-zk proof variants
• Governance parameters extended with PID tuning knobs

Testing

• cargo test --workspace passes
• cargo clippy --all-targets -- -D warnings clean
• cargo fmt --all -- --check passes
• no_std verification for all crates
• PID controller step response and stability tests
• BFT health consensus with simulated Byzantine nodes
• RS encode/decode roundtrip with erasure recovery
• KZG commitment/opening/batch-verify correctness
• PID + BFT integration test under load scenarios
• RS storage integration with epoch lifecycle
• Regression: no performance degradation in happy-path storage reads

Milestone

v0.9.3 - Orchestration Intelligence (due: 2026-09-30)

Checklist

• cargo fmt
• cargo clippy
• cargo test
• no_std check
• Documentation updated
• CHANGELOG.md updated
• KZG trusted setup documentation
• PID tuning guide for operators