Taking the next step in mainframe observability

Despite their frequently less visible role, mainframes remain indispensable in supporting the critical operations of many large organisations. With technology advancements over recent years, however, the mainframe has become an increasingly more versatile environment. The latest generation of IBM Z machines are capable of running cloud native, AI and modern analytics workloads with full hardware-level optimisation.

This has led to an emerging new role for the mainframe as a high scale hybrid cloud server. The extreme levels of scalability, resilience and security offered by IBM Z generally match or exceed the capabilities of public cloud services, with the added advantage of execution in an on-premises data centre or private hosted environment that you totally control.

Consequences of the expanded role

While the direction of travel is clearly very positive, providing mainframe customers with an opportunity to leverage their IBM Z investments differently and more broadly, it does challenge the way in which some aspects of mainframe operations have traditionally been handled.

With a more diverse and dynamic mix of workloads, it has become ever more difficult for operators to monitor and manage the environment effectively using the toolset they’ve historically relied on. While many of the specialist tools in place are great for what they were designed to do, lack of integration has got in the way of maintaining a complete and coherent view of overall system health. Not so much of a problem in the past when the pace of change was slower, but it’s become more challenging as Agile and DevOps based delivery have crossed over into the mainframe space. Add to this the inherently interconnected nature of many modern workloads, and the frequent need to maintain proximity between data and applications (e.g. in areas such as AI), and coherent observability can be hard to achieve, which in turn hampers effective service management.

Beyond the IBM Z itself, there’s also the need for observability, troubleshooting and remediation across boundaries when the mainframe is being exploited in a hybrid-cloud context. Whether it’s working in tandem with the public cloud or non-mainframe on-premise stacks, site reliability engineers (SREs) and others responsible for the performance and reliability of distributed applications need visibility across all of the components. In many cases, the individuals and teams involved will not have in-depth mainframe knowledge, and will be working day-to-day with their own tool-chains.

These evolving challenges and requirements form the backdrop for a recent announcement by the Mainframe Software business of Broadcom.

Joining the dots, including across boundaries

Broadcom’s new WatchTower Platform aims to address the observability challenges outlined above by providing a holistic, easy-to-use solution that caters for the needs of both traditional mainframe operators and SREs.

As Nicole Fagen, who leads AI Ops and Automation at Broadcom, explained during a recent briefing, Watchtower brings together capabilities that enable mainframe operators at different levels to quickly identify issues and their business impact. The solution provides a unified view that integrates alerts and contextual information, machine learning-driven anomaly detection, and application topology mapping. This allows Level 1 operators to resolve incidents quickly without having to traverse multiple tools and systems, and enables subject matter experts to drill down into specific resource-related views as necessary.

Beyond this, the inclusion of technology from Broadcom’s recent Z/IRIS (Integratable Real-time Information Streaming) acquisition extends end-to-end observability beyond the mainframe. By feeding mainframe operational data into distributed monitoring solutions such as Splunk, Datadog, Dynatrace and AppDynamics, Watchtower allows SREs to observe composite application behaviour and performance in a joined-up manner using tools with which they are already familiar.

The future of the mainframe in the enterprise

With the latest IBM Z technology enabling the mainframe to become a powerful hybrid-cloud platform, solutions like Watchtower that provide comprehensive end-to-end observability are key to overcoming traditional operational constraints. By empowering both mainframe and distributed teams, and enabling closer collaboration between them, these types of solutions allow organisations to embed the mainframe as an integral part of modern hybrid-cloud delivery.

Implemented appropriately, this creates a win-win situation in which the enterprise benefits from being able to exploit the full range of attributes of each platform as part of a coherent hybrid architecture. The scalability, resilience, security and data-processing power of the mainframe can all be brought to bear in a flexible and integrated manner alongside the rapid access to innovation enabled by the public cloud.

Data Center
Data Management