After a longer JobScheduler Master Pause orders take a long time to start

Current Situation

• A user runs some 14 000 job executions per day (observed for release 1.11.5)
• If a Master is paused and continued for 30 minutes then orders start properly.
• If a Master is paused and continued for some hours (with maybe some 1000 orders or more waiting in the queue) then it will take hours for all orders to start

Desired Behavior

• Orders should start immediately after a pause, independently from the duration of that pause.

Maintainer Notes
This issue is dismissed as we cannot reproduce the issue. JobScheduler starts orders immediatly, no matter paused for 30 minutes or 1/2/3/4/5 hours.

Test setup 1

• Single instance JobScheduler Release linux-x64 1.11.5
• Ubuntu 16_10

Test setup 2

• Passive cluster , As primary Instance - JobScheduler Release linux-x64 1.11.5
• Ubuntu 16_10

Test steps

1. Single Job Chain with two jobs and max_orders=3
   • job 01_TaskA1 sleeps for 2 seconds
   • job 01_TaskA2 sleeps for 1 second
2. Generated 1500 permanent orders, all scheduled to start at the specific time in future e.g. 17:30 PM
3. Once orders start, let first 50-100 orders process and then pause the JobScheduler.
4. Wait for 5 minutes so that JobScheduler updates the database with the orders' status.
5. Change the system date 1/2/3/4/5 hours in the future.
6. Wait for 5 minutes and continue the JobScheduler.

Test results

• The orders are started immediately, irrespective of pause duration.
• The behaviour is consistent for the Ad hoc orders and permanent orders.