The Shadow Scheduler did not take over after reboot
Article ID: 144411
Updated On:
Products
CA Workload Automation AE - Scheduler (AutoSys)
Autosys Workload Automation
Issue/Introduction
We stopped the Primary Server for maintenance. We expected an automatic failover to the Secondary Scheduler.
But after the Primary Server was rebooted, the Shadow Scheduler did not take over.
Environment
Workload Automation AE (AutoSys)
Resolution
The Shadow Scheduler takes over automatically when there is a brutal shutdown of the Primary Server because of a network interruption or a Server down issue.
In this case the AutoSys Services were stopped properly and manually before the reboot, that is why the shadow did not take over.
When you have a maintenance operation planned on the Primary or the Shadow or TB node in a "High Availability" environment, the best solution is to proceed by the following method:
1. Stop all the AutoSys Services (Application Server, Scheduler and Agent) on the Primary and Shadow nodes (and TB node if Full HA with DUAL Databases)
2. Edit the config.$AUTOSERV file under $AUTOUSER directory to configure the node which will stay up in Single Server mode
3. Restart AutoSys Services (Application Server, Scheduler and Agent) Only on the node which will stay up during the maintenance operation of the other node
4. When the maintenance operation is completed, stop again all the AutoSys Services (Application Server, Scheduler and Agent) on the Primary and Shadow nodes (and TB node if Full HA + Dual DB)
5. In case you are in "Full HA" mode with "Dual Server", make sure databases have remained synchronized. Otherwise synchronize databases by "autobcp" before restarting in Full HA mode
6. Update the config.$AUTOSERV files on the Primary, Shadow and TB nodes to configure them in (Full) HA mode
7. Restart all the AutoSys Services (Application Server, Scheduler and Agent) on the Primary, Shadow and TB nodes
In this case the AutoSys Services were stopped properly and manually before the reboot, that is why the shadow did not take over.
When you have a maintenance operation planned on the Primary or the Shadow or TB node in a "High Availability" environment, the best solution is to proceed by the following method:
1. Stop all the AutoSys Services (Application Server, Scheduler and Agent) on the Primary and Shadow nodes (and TB node if Full HA with DUAL Databases)
2. Edit the config.$AUTOSERV file under $AUTOUSER directory to configure the node which will stay up in Single Server mode
3. Restart AutoSys Services (Application Server, Scheduler and Agent) Only on the node which will stay up during the maintenance operation of the other node
4. When the maintenance operation is completed, stop again all the AutoSys Services (Application Server, Scheduler and Agent) on the Primary and Shadow nodes (and TB node if Full HA + Dual DB)
5. In case you are in "Full HA" mode with "Dual Server", make sure databases have remained synchronized. Otherwise synchronize databases by "autobcp" before restarting in Full HA mode
6. Update the config.$AUTOSERV files on the Primary, Shadow and TB nodes to configure them in (Full) HA mode
7. Restart all the AutoSys Services (Application Server, Scheduler and Agent) on the Primary, Shadow and TB nodes
Feedback
Yes
No
Powered by
