...

• Users might be interested in specifying jobs and dependencies from Microsoft Excel® spreadsheets.
  • In a situation when business staff is in charge of specifying jobs then Excel® is a frequently used tool.
  • As a no-code approach this allows to specify jobs and dependencies by users without detailed know-how of JS7 - Workflows.
• The solution explained from the article creates jobs and workflows from Excel® spreadsheets ready for import into JS7.
  
  • The solution makes use of a PowerShell cmdlet for Linux and Windows and can be operated for PowerShell 5.1, 6, 7.
  • The solution works for .xlsx files. Files from formats such as .xls and .csv have to be imported to Excel®.
  • The solution is both a tool for simple use cases and a programming example how to create workflows from Excel®. More elaborate approaches for processing of customized Excel® input are subject to SOS professional services.

Prerequisites

The following PowerShell modules are used for this purpose:

...

• The worksheets included with the spreadsheet file offer a number of jobs and dependencies from the following columns:
  • NodeID: A unique identifier for the occurrence of a job. If the same job occurs a number of times then the Name column holds the same job name, however, a unique value has to be specified for the NodeID column.
  • SuccessorID: Specifies the NodeIDs of jobs that are direct successors of the current job. If more than one job is specified then they will will be executed in parallel branches. A number of NodeIDs in this column are separated by a space.
  • Instruction:  Makes use of the fixed value Job. This column is reserved for later support of further JS7 - Workflow Instructions.
  • Name: Specifies the job name.
  • Description: Optionally specifies a descriptive title for the job.
  • Agent, Subagent Cluster: Specifies the name of a Standalone Agent that is configured in your JS7 scheduling environment. If an Agent Cluster is used then its name is specified as the Agent and the respective Subagent Cluster is specified, see JS7 - Agent Cluster.
  • Script: Specifies the job script that will be executed. Multiline input is allowed.
  • Error Handling: Specifies Optionally specifies one of the following options for for JS7 - How to apply error handling:
    • STOP: If the job fails then the order will be put to the FAILED state and will remain with the job. Failed orders require user intervention to resume, to suspend or to cancel the respective order.
    • IGNORE: Any job error is ignored and order processing continues, see JS7 - Try-Catch Instruction.
    • RETRY: The job will be retried in case of error for 3 times with a delay of 60s, see JS7 - Retry Instruction.
    • LEAVE: The order will leave the workflow with an unsuccessful History outcome, see JS7 - Finish Instruction.
  • Fail on stderr: Specifies Optionally specifies that in case that the job will write output to the stderr channel the job will be considered to be failed, see see JS7 - Job Instruction.
• Users can adjust column names, see JS7 - PowerShell Import from Excel Script - New-JS7WorkflowFromExcel.ps1.

...

• Job dependencies are designed along a Directed acyclic graph (DAG). This includes a few rules for input to the Excel® spreadsheet:
  • One or more initial jobs are required as the graph's root node(s). Initial jobs must not be specified as successors in any of the SuccessorID column values.
  • One A final job has to can be specified that does not indicate a successor job in the its SuccessorID column.
  • The order of entries to the Excel® spreadsheet is not relevant.
  • Not too much error checking of the DAG is performed, however, missing job nodes will be detected.
• In a simplified way a directed graph works with represents job dependencies from two columns:
  • The NodeID column specifies the unique identifier of the given occurrence of a job.
  • The SuccessorID column specifies one or more NodeIDs of successor jobs that are separated by a space.
    • If more than one NodeID is specified then this will fork processing of the indicated successor jobs to be performed in parallel branches.
    • When all To join parallel branches of jobs the NodeIDs specified with the from the SuccessorID column of a given job node indicate the same NodeID as the SuccessorID with their individual entries then they will be joined before the indicated successor job nodethe forking job have to hold empty values in their respective SuccessorID column or hold the same NodeID of a later job.
    • Forking can include any number of parallel branches specified by NodeIDs in a SuccessorID column.
    • Forking can be nested preferably at not more than 15 levels.

...