Skip to main content

Decentralized Orchestration

Deeploy-first orchestration across licensed edge infrastructure

Decentralized Orchestration

Ratio1 orchestration controls where and how workloads run across licensed edge infrastructure, including rollout policy, resource intent, and settlement-aware execution.

Production-first orchestration model

For production services, the primary path is Deeploy on licensed node capacity, through:

  • Deeploy App (deeploy.ratio1.ai) for managed UI operations.
  • Deeploy API for programmatic production orchestration.
  1. Define deployment intent (service type, runtime model, resources, policies).
  2. Select eligible node capacity through orchestration workflows.
  3. Launch and monitor execution through Deeploy UI/API surfaces.
  4. Operate updates and scaling through the same lifecycle channel.
  5. Follow oracle-verified, escrow-based economics for funded execution and settlement.

This reduces direct node-by-node operational burden for teams running real services.

Oracle verification and settlement path

Production jobs are funded through escrow and observed through oracle/runtime telemetry signals. Settlement is conditional: when uptime/execution conditions are met for the relevant windows, rewards can be released to participating roles and nodes.

When conditions fail, orchestration can reallocate workloads, while trust/economic controls can apply withholding, penalty, or burn-linked outcomes depending on policy.

CAR and WAR execution paths

Deeploy commonly exposes two app delivery patterns:

  • CAR (Container App Runner): container-image oriented deployment.
  • WAR (Worker App Runner): Git-to-edge worker app flow with build/run lifecycle management.

Both rely on the same decentralized plugin-based runtime, while differing in packaging and delivery workflow.

ChainDist and direct SDK orchestration

Ratio1 also exposes native orchestration paths outside managed app-style deployment:

  • ChainDist for cooperative distributed jobs (fan-out workers + aggregation), with smart-contract-driven job lifecycle signals (for example job ID/payment checks, oracle validation updates, and close/reward transitions).
  • Direct SDK orchestration for explicit pipeline/instance control, custom control-plane logic, and advanced integration workflows.

Deeploy, ChainDist, and direct SDK orchestration are complementary layers on the same runtime. For most production fleets, Deeploy is the default control plane; ChainDist and direct SDK are used when teams need lower-level distributed or programmatic control.

Role split in orchestration

  • Node Operator: provides licensed compute capacity and node uptime.
  • CSP: manages service deployment and lifecycle via Deeploy workflows.
  • Developer: builds workload logic and integrations against runtime/data interfaces.

CSP production onboarding includes oracle-aligned deed/license requirements plus KYC/KYB checks. Trust protocol governance provides freeze/suspend/blacklist enforcement paths for serious violations.

Ground truth references

Primary:

Supporting:

Notable date

  • Reviewed on February 18, 2026.

Next steps