Validator Responsibilities

Validators are responsible for more than uptime. They are part of the replicated state machine, so operational discipline directly affects consensus.

Core Responsibilities

Validators are expected to:

• run the supported software versions for the network they joined
• protect validator and operator key material
• keep the node online, healthy, and correctly peered
• monitor height, sync status, mempool health, and runtime metrics
• apply approved network upgrades and recovery procedures in a coordinated way

Determinism Discipline

The current supported validator runtime is the fixed xian_vm_v1 path.

Validators must stay aligned on:

• xian-abci and xian-contracting
• the native VM runtime capability
• the fixed native VM gas schedule
• canonical Rust source-to-IR compilation for submitted contracts

Ad hoc local runtime changes are dangerous because validators replay the same deterministic VM state machine.

Operational Safety

Good validator posture includes:

• testing changes in localnet or smoke flows before rollout
• keeping snapshots, state-sync inputs, and patch bundles organized
• exposing only the ports your deployment model actually needs
• separating validator duties from optional BDS and sidecar extras when necessary
• watching for mismatches in execution, metrics, and indexed-service recovery

What Validators Should Not Do

Validators should not:

• patch consensus-sensitive runtime code independently
• drift onto uncoordinated runtime versions
• improvise execution-policy settings on a live network
• treat optional dashboards, BDS, or relayers as if they were part of consensus-critical state

The safest posture is boring consistency.