THE config.json FILE

The most important part of using ACT is the use of the configuration file. This file must contain all the details of the tasks that need to be executed. This will provide uncomplicated customization and access to most features of the Microsoft Azure Batch Services.

This file must contain a few required configuration strings and many optional(*) ones. Below is a description of the current configuration strings of this tool:

Configuration attribute	Type	Description
batch	---	The batch account configuration
batch.accountName	string	The name of the batch account on your Microsoft Azure Subscription to be used
batch.accountKey	string	The corresponding batch account key, found in the Azure portal
batch.accountUrl	string	The corresponding batch account URL, found in the Azure portal
pool	---	These configuration parameters come from the PoolAddParameter Class
pool.id	string	The ID that will be used to name the pool of compute nodes created
pool.dedicatedNodeCount	integer	The number of dedicated compute nodes to be created in this pool
pool.lowPriorityNodeCount	integer	The number of low-priority compute nodes to be created in this pool
pool.vmSize	string	The string representing the VM size to be used
pool.vmConfiguration	---	The Virtual Machine (VM) configurations
pool.vmConfiguration.imageReference	---	The Virtual Machine image configuration to be used
pool.vmConfiguration.imageReference.publisher	string	The VM Image publisher string
pool.vmConfiguration.imageReference.offer	string	The VM Image offer description string
pool.vmConfiguration.imageReference.sku	string	The VM Image operation system string
pool.vmConfiguration.imageReference.version	string	The VM Image version string
pool.vmConfiguration.nodeAgentSKUId	string	The VM Image Agent SKU ID to be used
pool.useEphemeralOSDisk	bool	If the VM should use ephemeral OS Disks, can only be used if there are only dedicated nodes
pool.nodeStorageContainers	---	Used to mount storage containers in the compute nodes. It’s important to note that if different Tasks use the same blob in the same container, this can cause synchronizing problems that have to be considered, for more details see the blobfusion specifications.
pool.nodeStorageContainers.mount	bool	Specifies if the nodes should mount the following storage containers
pool.nodeStorageContainers.containers*	vector	A list with all the containers to be mounted
pool.nodeStorageContainers.containers.name*	string	The storage container name
pool.nodeStorageContainers.containers.blobfuseOptions*	string	The blobfuse options to the container mount procedure
pool.nodeAutoScale	---	Used to include auto scale property to the pool
pool.nodeAutoScale.include	bool	Specifies if the node should use an auto-scale formula
pool.nodeAutoScale.evaluationInterval*	integer	The autoscale evaluation interval
pool.nodeAutoScale.formula*	vector [string]	A list of strings representing the autoscale formula
pool.applications	---	Used to include applications on the compute nodes
pool.applications.include	bool	Specifies if the compute nodes should install the applications
pool.applications.references*	vector	A list of references to the applications to be installed
pool.applications.references.id*	string	The application ID
pool.applications.references.version*	string	The application version
pool.startupTask	---	Used to include a Startup Task to the compute nodes
pool.startupTask.include	bool	Specifies if the compute nodes should include a startup task
pool.startupTask.command*	string	The command line of the startup task
job	---	Used to include the job configurations
job.id	string	The Job ID
tasks	---	Used to include the tasks configurations from TaskAddParameter Class
tasks.addCollectionStep	integer	The number of tasks to be included in each iteration
tasks.inputs	---	The definitions of where and how to gather inputs to be used
tasks.inputs.areBlobsInInputStorage	bool	Specifies if the inputs are to be collected from the input storage, if FALSE an input file should be provided in the execution arguments
tasks.inputs.inputFileExtension*	string	The file extension of the blobs to be used as inputs
tasks.inputs.outputFileExtension*	string	The file extension expected to be on the output blobs
tasks.inputs.filterOutExistingBlobInOutputStorage*	bool	Specifies if should remove from the input list the blobs that already have their correspondent output blob in the output storage
tasks.inputs.filterOutExistingTaskInCurrentJob	bool	Specifies if should remove from the input list the blobs that already have a Task assigned with the same input
tasks.inputs.requiredSlotFormula	vector [string]	Used to assign a specific formula to calculate the required task slot of each Task. This formula should be written using Python syntax, and can read any configuration parameter using $ before its name and a dot for each hierarchical level like '$pool.dedicatedNodeCount'. Has as built-in parameters inputName and inputSize
tasks.inputs.order*	---	Used to define the sorting order of the input list
tasks.inputs.order.by*	string	Attribute used to sort the list. Possible values are: name, size, or slot count. The default value is to order by name
tasks.inputs.order.type*	string	Used to define the sorting method, possible values are asc or desc, the default value is asc
tasks.resources	---	Used to define if resources should be copied to the Task node automatically
tasks.resources.automaticInputsUpload	bool	Define if it should copy the input blobs to the Task working directory
tasks.resources.automaticScriptsUpload	bool	Define if it should copy the script blobs to the Task working directory
tasks.logs	---	Used to define what to do with log files after the Task goes to completed state
tasks.logs.automaticUpload	bool	Define if logs are to be copied out of the Tasks when they finish
tasks.logs.destinationPath*	string	Define the destination path where to copy the logs from Tasks when they finish
tasks.outputs	---	Used to define what to do with the output after the Task goes to successfully completed state
tasks.outputs.automaticUpload	bool	Define if output blobs are to be copied out of the Tasks when they finish
tasks.command	string	Define the Task commandLine. For each Task the command executed will be appended with its received input, enclosed in single quotes, followed by the tasks.commandSuffix parameter if it exists.
tasks.commandSuffix*	string	Define the end of the Task commandLine. For each Task the command executed will be appended with its received input, enclosed in single quotes, followed by the tasks.commandSuffix parameter if it exists. Default value is an empty string ("")
tasks.retryCount	integer	Define the retry count of the Tasks
tasks.retentionTimeInMinutes	integer	Define the time in minutes to retain the Task files in the compute node
storage	---	The storage account configurations
storage.accountName	string	The name of the storage account you are going to use
storage.accountDomain	string	The storage account domain you are going to use
storage.accountSASToken	string	The storage account SAS Token you are going to use
storage.scripts	---	Used to define where are the scripts in the storage
storage.scripts.container	string	Define the container name where are the scripts
storage.scripts.blobPrefix	string	Define the blobPrefix of the scripts in the specified storage container
storage.input	---	Used to define where are the inputs in the storage
storage.input.container	string	Define the container name where are the inputs
storage.input.path	string	Define the path of the inputs in the specified storage container
storage.input.blobPrefix	string	Define the blobPrefix of the inputs in the specified storage container
storage.output	---	Used to define where are the outputs in the storage
storage.output.container	string	Define the container name where are the outputs
storage.output.path	string	Define the path of the outputs in the specified storage container
storage.output.blobPrefix	string	Define the blobPrefix of the outputs in the specified storage container
cleanup	---	Used when call the free resources argument
cleanup.timeoutInMinutes	integer	Defines the time-out in minutes for the cleanup process. Terminate the process after this time, if it does not finish before that

* These parameters are optional if the include parameter that comes before them is FALSE.