High availability with IBM z/OS Connect

A high availability (HA) solution helps minimize disruption to daily operations caused by the failure of one or more components within the overall system. This section outlines the core principles of high availability and how they can be applied in the context of z/OS Connect to build resilient API infrastructure.

This information is intended for architects and system programmers responsible for implementing a high availability (HA) environment for z/OS Connect. Successful implementation requires system programming expertise, including configuring TCP/IP workload balancing solutions and creating zFS file systems to support shared resources.

There are three main areas to consider high availability for z/OS Connect APIs:
  • Client Connectivity - Clients, including API gateways, should be configured to use persistent TLS connections to z/OS Connect Servers. This reduces CPU overhead on both the client and server sides. Connections should target shared ports to enable load distribution across multiple z/OS Connect Server instances.
  • z/OS Connect Server Deployment - Deploy multiple z/OS Connect Servers to eliminate single points of failure. This can ensure that API availability is maintained even if one server becomes unavailable.
  • System of Record Resilience - Backend systems of record such as CICS, Db2, IMS, and IBM MQ must also be configured for high availability. This includes the use of clustering, workload balancing, or other failover mechanisms to help ensure continuous service and avoid bottlenecks or outages.

It’s essential to plan for workload variability, particularly during periods of increased demand when extra system capacity might be required. The system must respond predictably to fluctuations in workload to consistently meet service level agreements (SLAs). For z/OS Connect, scalability is achieved by increasing the number of server instances, allowing the environment to handle higher volumes of API traffic without compromising performance or availability.

Containers High availability for z/OS Connect Server images can be achieved by using the scaling and orchestration capabilities of your container platform such as Red Hat® OpenShift® or Kubernetes. By deploying multiple z/OS Connect instances across nodes, you can ensure resilience, load distribution, and failover protection. These platforms support automated scaling, health monitoring, and restart policies, which help maintain service continuity in the event of node or container failure. For more information, see IBM z/OS Connect Server image high availability.

Started task You can integrate various IBM and third-party products with z/OS Connect to build a high availability (HA) solution that is tailored to your organization’s needs. Deployment options range from a simple configuration with two z/OS Connect Servers running in a single LPAR, to a more advanced parallel sysplex architecture with multiple z/OS Connect Servers distributed across several LPARs. This flexibility allows you to design a solution that aligns with your scalability, resilience, and operational requirements.