Workflow Options Overview

Workflow options can affect the execution of a single workflow without having to change configuration options or restart Cromwell.

You provide workflow options to Cromwell in a JSON format. This can be supplied at workflow-submit time either via the CLI or the REST endpoint:

    "option_name_1": "option value 1",
    "option_name_2": "option value 2"

Unless otherwise specified you can expect workflow options to override any hard-coded defaults in Cromwell or defaults provided in the configuration file, but to be overridden by any values provided in the workflow definition file itself (WDL).

Some workflow options apply only to tasks running on the Google Pipelines API backend.

Global Workflow Options

The following workflow options apply to all workflows and their calls regardless of the backend being used.

Runtime Attributes

Some options allow you to override or set defaults for runtime attributes.

Setting Default Runtime Attributes

You can supply a default for any Runtime Attributes by adding a default_runtime_attributes map to your workflow options file. Use the key to provide the attribute name and the value to supply the default.

These defaults replace any defaults in the Cromwell configuration file but are themselves replaced by any values explicitly provided by the task in the WDL file.

Example options.json:

    "default_runtime_attributes": {
        "docker": "ubuntu:latest",
        "continueOnReturnCode": [4, 8, 15, 16, 23, 42]

In this example, if a task in a workflow specifies a docker: attribute, the task will get what it specifies. However if any task does not provide a value then it will be treated as though it had specified ubuntu:latest.

Specific Runtime Attributes

Option Value Description
continueOnReturnCode true or false or integer array Globally overrides the continueOnReturnCode runtime attribute for all tasks
backend An available backend Set the default backend specified in the Cromwell configuration for this workflow only.

Example options.json:

    "continueOnReturnCode": false,
    "backend": "Local"

In this example, all tasks will be given to the Local backend unless they provide a value explicitly in their runtime { ... } block. In addition, the continueOnReturnCode value for all tasks is hard-coded to false, regardless of what the tasks put in their runtime block. TODO or is just a default ala default_runtime_attributes?

Workflow Failure

The workflow_failure_mode option can be given the following values. This overrides any default set by the workflow-options.workflow-failure-mode configuration options.

Value Description
ContinueWhilePossible Continues to start and process calls in the workflow, as long as they did not depend on the failing call.
NoNewCalls No new calls are started but existing calls are allowed to finish.

Example options.json:

    "workflow_failure_mode": "ContinueWhilePossible"

Output Copying

Option Value Description
final_workflow_outputs_dir A directory available to Cromwell Specifies a path where final workflow outputs will be written. If this is not specified, workflow outputs will not be copied out of the Cromwell workflow execution directory/path.
use_relative_output_paths A boolean When set to true this will copy all the outputs relative to their execution directory. my_final_workflow_outputs_dir/~~MyWorkflow/af76876d8-6e8768fa/call-MyTask/execution/~~output_of_interest . Cromwell will throw an exception when this leads to collisions. When the option is not set it will default to false.
final_workflow_log_dir A directory available to Cromwell Specifies a path where per-workflow logs will be written. If this is not specified, per-workflow logs will not be copied out of the Cromwell workflow log temporary directory/path before they are deleted.
final_call_logs_dir A directory available to Cromwell Specifies a path where final call logs will be written. If this is not specified, call logs will not be copied out of the Cromwell workflow execution directory/path.

Note that these directories should be using the same filesystem as the workflow. Eg if you run on Google's PAPI, you should provide gs://... paths.

Example options.json:

    "final_workflow_outputs_dir": "/Users/michael_scott/cromwell/outputs",
    "use_relative_output_paths": true,
    "final_workflow_log_dir": "/Users/michael_scott/cromwell/wf_logs",
    "final_call_logs_dir": "/Users/michael_scott/cromwell/call_logs"

With "use_relative_output_paths": false (the default) the outputs will look like this


The above result will look like this when "use_relative_output_paths": true:


This will create file collisions in final_workflow_outputs_dir when a workflow is run twice. When cromwell detects file collisions it will throw an error and report the workflow as failed.

Call Caching Options

These options can override Cromwell's configured call caching behavior for a single workflow. See the Call Caching section for more details and how to set defaults. The call caching section will also explain how these options interact when, for example, one is set true and the other is false.

Note: If call caching is disabled, these options will be ignored and the options will be treated as though they were false.

Option Values Description
write_to_cache true or false If false, the completed calls from this workflow will not be added to the cache. See the Call Caching section for more details.
read_from_cache true or false If false, Cromwell will not search the cache when invoking a call (i.e. every call will be executed unconditionally). See the Call Caching section for more details.

Example options.json:

    "write_to_cache": true,
    "read_from_cache": true

Retry with More Memory Multiplier

The memory_retry_multiplier workflow option sets the factor by which the memory should be multiplied while retrying when Cromwell encounters one of the error keys (specified in Cromwell config using system.memory-retry-error-keys) in the stderr file. The factor should be in the range 1.0 ≤ multipler ≤ 99.0 (note: if set to 1.0 the task will retry with same amount of memory). If this is not specified, Cromwell will retry the task (if applicable) but not change the memory amount. See the Retry with More Memory section for more details.

Example options.json:

    "memory_retry_multiplier" : 1.1