Check orders from a daily plan for latecomers

Current Situation

• Users configure jobs to launch ad hoc orders on a daily basis.
• Such orders are not permanent but are expected to be completed and removed within the period of the daily plan.
• If such orders cannot be completed, e.g. due to stopped jobs or waiting sync nodes in a job chain, then they will remain in the system up to the point in time when the affected job is unstopped or the sync conditions are met. This usually happens when an order with the same scope arrives and might result in the fact that the current order and the previous order (latecomer) would both continue in a job chain with unwanted results.

Desired Behavior

• A job should check the system for latecomers which includes orders that exceed the lifetime of similar orders in the job history
  • by a configurable percentage or absolute value. A possible default is an increase of 30%.
  • The calculation is done based on the average lifetime of a configurable number of predecessor orders (default: 10).
• This job should support checking sync nodes, i.e. to show the dependency of a number of orders that are waiting at the same sync job.
• The output of this job is a plain text list of latecomer orders sorted by the exceeding values.
  • The output is added to the task log.
  • The output can be configured to be written to a location in the file system.
• This job can be configured to run
  • for specific folders in the live directory
  • for specific job chains
  • for specific exceeding values, e.g. orders exceeding 30% or more of the average lifetime
• This job allows the following actions
  • list: (default) create a list of latecomers
  • delete: removes latecomer orders from the system

1. Reports showing the duration of job chains and job chain nodes
The data base for the reports are the table REPORT_EXECUTIONS and REPORT_TRIGGERS

Reported values for nodes in job chains are

• maximum duration
• minimum duration
• average duration
• number of runs

Available reports

• per job chain node
  • Parameters: job chain, state, orderid
    • values for one specific orders with one specific order id are listed
    • the value * for order creates a report for all order seperated by ids
  • Parameters: job chain, state
    • values for one specific node are listed. This can incluce orders with different order ids.
    • the value * for the state creates a report for all nodes of the given job chain seperated by nodes
  • Parameters: job chain
    • values for one specific job chain are listed.
• List suspended orders
  • filtered by a job java class (default=com.sos.jitl.sync.JobSchedulerSynchronizeJobChainsJSAdapterClass)
  • option warn_if_execution_time_exceeds average execution time by warn_on_percentage_value amount (default=10).
• List running orders
  • these orders are identified by orders in SCHEDULER_ORDER_HISTORY that do not have an end time.
  • filtered by a job java class (default=com.sos.jitl.sync.JobSchedulerSynchronizeJobChainsJSAdapterClass)
  • option warn_if_execution_time_exceeds average execution time by warn_on_percentage_value amount (default=10).

Maintainer Notes
This feature is partially available from the JOC Cockpit - Daily Plan with release 1.11.