In the event of a SAS^® Grid Manager controller node failure, job flows that are scheduled in SAS Job Flow Scheduler are not passed to a new controller node so that they can run as scheduled.

However, when high availability is enabled, if one node fails, another node is available to continue running jobs. Using high availability will change the underlying communication method for SAS Job Flow Scheduler. Previously, SAS Job Flow Scheduler communicated between its scheduling server and the SAS^® Web Infrastructure Platform Database via TCP connections on ports 1964 and 1981. With high availability, SAS Job Flow Scheduler uses the Java Messaging Service (JMS) to queue requests between the two systems. 

Here is how the process works:

• SAS^® Workload Orchestrator fails to start the OS Trigger service for another node (or even the same node). Then the OS Trigger service performs these actions:

1. Registers its node as the primary node for SAS Job Flow Scheduler
2. Synchronizes the triggers of the scheduled flows to SAS Job Flow Scheduler on the current node, retrieving the trigger information from the WebDAV service in the middle tier
3. Starts listening for messages to schedule, unschedule, or reschedule flows

To take advantage of the new high availability feature for SAS Job Flow Scheduler, apply the hot fix that is provided in this SAS Note.

Click the Hot Fix tab in this note to access the hot fix for this issue.

Operating System and Release Information

A fix for this issue for SAS Workload Orchestrator and SAS Grid Manager for Platform 9.47 is available at:

A fix for this issue for SAS Middle Tier 9.4_M7 is available at:

A fix for this issue for SAS Job Flow 9.47 is available at:

A fix for this issue for SAS Management Console 9.4_M7 is available at: