ccc2e94c2a
The reference graph processed every work unit inside a single op, so exactly one worker was ever reported regardless of executor. That made the guide's claim that k8s_job_executor yields several distinct contributing_hosts false, and left the reference implementations unable to demonstrate the executor choice at all. generate_work_units is now a DynamicOut and both graphs map over it, so one step is created per unit and the loosely coupled pattern dispatches one external workload per unit. Evidence is split into contributing_workers (host and pid, differs per process) and contributing_hosts (differs only across machines), because the previous single field could not distinguish multiprocess fan-out from no fan-out. Tests now assert the mapped step keys rather than a host count, since execute_in_process ignores executor_def and cannot prove executor behaviour on its own. Adds a Windows note: multiprocess_executor did not complete during authoring and left orphaned processes. Changelog: fixed
3.4 KiB
Distributed Execution
Canonical location for documentation, example workflows and reference service implementations covering distributed execution patterns in the Simpl orchestration platform.
The service answers one question for every workflow: where does the compute actually run, and how does runtime information get back to the control plane?
Contents
| Area | Location | Status |
|---|---|---|
| User guide | documents/user-guide/distributed-execution-guide.md | Complete |
| Readiness checklist | documents/user-guide/readiness-checklist.md | Complete |
| Tightly coupled reference | src/distributed_execution/tightly_coupled/jobs.py | Runnable |
| Loosely coupled reference | src/distributed_execution/loosely_coupled/jobs.py | Runnable (subprocess verified) |
| External payload | payload/work.py | Runnable |
| Example configuration | yaml/ | Complete |
Project structure
distributed-execution/
├── src/
│ └── distributed_execution/
│ ├── repository.py # Dagster definitions (entry point)
│ ├── ops.py # Shared ops used by both patterns
│ ├── preflight.py # Readiness checks backing the checklist
│ ├── tightly_coupled/
│ │ └── jobs.py # in_process / multiprocess / k8s_job_executor
│ └── loosely_coupled/
│ └── jobs.py # PipesSubprocessClient / PipesK8sClient
├── payload/ # External workload: dagster-pipes ONLY
│ ├── work.py
│ ├── requirements.txt
│ └── Dockerfile
├── documents/user-guide/ # AC1-AC4 documentation
├── yaml/ # Working example configuration
├── tests/
├── Dockerfile
├── pyproject.toml
└── workspace.yaml
Getting started
Prerequisites: Python 3.12+ and uv.
uv sync --dev
uv run dagster dev -f src/distributed_execution/repository.py
The Dagster UI is then available at http://localhost:3000. Two jobs run end to
end on a laptop with no cluster: tightly_coupled_in_process_job and
loosely_coupled_subprocess_job. The Kubernetes variants of each require a
cluster and are documented in the user guide.
tightly_coupled_local_jobusesmultiprocess_executor. It did not complete on a Windows development machine during authoring — see the Windows note in the user guide's section 5.5.
Running tests
uv run pytest
Building the images
Two images, deliberately: the code location and the external payload are versioned and scanned independently.
docker build -t distributed-execution:0.1.0 .
docker build -f payload/Dockerfile -t distributed-execution-payload:0.1.0 payload/
Status
Both execution targets are implemented. The tightly coupled jobs and the loosely coupled subprocess transport are verified by the test suite. The loosely coupled Kubernetes transport is implemented but has not yet been run against a cluster; see the guide's Outstanding work section.
Licence
European Union Public Licence v1.2 — see LICENSE.