IBM Multi-site Workload Lifeline, or Lifeline, is a product that has the ability to monitor the applications that comprise a customer’s business-critical workloads across two z/OS active sysplexes, as well as to supply recommendations for how to route workload connections or MQ messages that target these applications.

Lifeline is also an integral part of GDPS Continuous Availability, GDPS CA, a solution that provides near continuous availability for z/OS workloads. GDPS CA accomplishes this by keeping workload data sources in sync across two sites, using software replication, and by monitoring and switching workloads, using Lifeline, to the alternate site in an event of a failure within the production site, typically in seconds.

Lifeline provides the routing recommendations to influence how external load balancers route workload connections or influence how MQ clusters distribute workload MQ messages. There is still a need for customers who have invested resources into their own workload routing infrastructure to be able to leverage their existing workload routing when deploying the GDPS Continuous Availability solution in their environment.

Lifeline user exits

With apar PH25729, Lifeline enables customers to add their own custom routing into the workload routing decision. Workloads can take advantage of the flexibility provided by Lifeline user exits for the following scenarios:

• Using existing monitoring tools to complement the monitoring performed by Lifeline when making workload routing decisions
• Using custom workload routing processes in place of the workload routing enabled by Lifeline

User exit support is enabled for a workload by configuring the workload in the new user_exit_list statement.

For each workload configured to utilize user exit processing, there is an option to replace Lifeline’s workload routing recommendations with custom routing, via the BYPASS keyword, or to complement Lifeline’s workload monitoring, via the NOBYPASS keyword.  When BYPASS is configured, Lifeline skips all its routing recommendations for the workload, and expects that the call to the user exit will handle the workload routing, itself. Optionally, there is a timeout keyword that tells Lifeline how long to wait for the user exit to complete its processing before Lifeline aborts the command being processed.
 

When a workload is configured to use a user exit, whenever a Lifeline command is issued to ACTIVATE, QUIESCE, or DEACTIVATE a workload, the user exit is called with the command being processed. The user exit can then perform additional monitoring of the workload  and/or handle the routing of the workload, before control returns back to Lifeline to continue its processing of the command.

Lifeline uses the z/OS Dynamic Exits Facility to provide the interface for loading, managing, and calling a user exit.

Lifeline ships sample Assembler and Metal C user exits, to provide customers a model for tailoring one for their own workload requirements.

Learn More

To learn more about IBM Multi-site Workload Lifeline and its user exit capabilities, check out https://www.ibm.com/products/multisite-workload-lifeline.

For information about the GDPS Continuous Availability solution, refer to https://www.ibm.com/it-infrastructure/z/technologies/gdps.

And to better understand the z/OS Dynamic Exits Facility, see https://www.ibm.com/support/knowledgecenter/en/SSLTBW_2.4.0/com.ibm.zos.v2r4.ieae400/dynexit.htm.

 
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Author: Michael Fitzpatrick

Michael Fitzpatrick is a Senior Technical Staff Member of the IBM Enterprise Networking Software Group, based in Research Triangle Park, North Carolina, in the US. He is the architect for the Multi-site Workload Lifeline product. Mike has worked in the networking area for 23 years, with a focus on resiliency, network design, and performance.