Skip to main content
Workflow orchestration tools define pipeline structure, dependencies, scheduling, retries, and execution state. They determine what should run and when, but they do not observe how tasks behave while they execute. OpenSRE complements workflow orchestrators by observing execution behavior at runtime. It captures how tasks and subprocesses consume CPU, memory, disk, and network resources, and maps this behavior back to pipeline runs, steps, and workflows.
If you’re new to OpenSRE or want a high-level mental model, see How OpenSRE fits in your stack.

​
Supported workflow orchestrators

The integrations below describe how OpenSRE works with common orchestration frameworks used in scientific, data, and bioinformatics pipelines. Each page focuses on the execution gaps specific to that tool.

OpenSRE and Apache Airflow

What DAGs don’t show at runtimeOpenSRE observes execution behavior inside Airflow tasks, including subprocesses and external tools.

OpenSRE and Dagster

Execution insight beneath assets and jobsOpenSRE reveals how resources are consumed during execution of Dagster assets, ops, and jobs.

OpenSRE and Flyte

Execution insight beneath tasks and workflowsOpenSRE shows how Flyte tasks actually behave at runtime, beyond task state and execution metadata.

OpenSRE and Prefect

Runtime visibility beneath flow stateOpenSRE captures system-level behavior of Prefect tasks, including work performed outside Python.

OpenSRE and Seqera

Execution behavior inside Nextflow pipelinesOpenSRE observes what happens inside Nextflow tasks at runtime, beyond scheduling and task state.

​
When OpenSRE is useful with orchestration tools

OpenSRE is most useful alongside workflow orchestrators when teams need to:
  • Understand why tasks run slower than expected
  • Distinguish CPU-bound, I/O-bound, and idle execution
  • Diagnose performance issues not visible in task logs
  • Attribute resource usage and cost to specific workflows or runs
OpenSRE does not replace orchestration. It does not change scheduling, retries, or pipeline definitions. It adds execution-level visibility beneath existing workflows.

​
Where to go next