Research Finder
Find by Keyword
A New Era for Mainframe: Seamless Integration via OpenTelemetry
Z Observability Connect 7.1.0 delivers OpenTelemetry traces, metrics, and logs; competition is fierce; hybrid cloud focus is paramount.
19/11/2025
Key Highlights:
The introduction of OpenTelemetry support aims to bridge the longstanding visibility gap between mainframe and distributed environments.
IBM Z Observability Connect 7.1.0 centralizes telemetry signals—traces, metrics, and logs—from critical z/OS subsystems.
A new Telemetry Controller, a custom OpenTelemetry Collector, is designed to transform mainframe data into vendor-agnostic OTLP format.
This move addresses the significant market trend toward hybrid observability and cross-platform Site Reliability Engineering practices.
Major competitors are also actively pushing OpenTelemetry integration, making this a fight for the standardization layer of the mainframe.
The News
IBM announced IBM Z Observability Connect 7.1.0, a significant update focusing on adopting the OpenTelemetry (OTel) standard across IBM Z environments. The new release aims to deliver comprehensive telemetry data, including traces, metrics, and logs, from z/OS systems like CICS, IMS, and Db2. This functionality allows IT operations teams to integrate mainframe performance insights into modern, enterprise-wide observability platforms. Find out more by clicking here to read the announcement blog.
Analyst Take
OpenTelemetry (OTel) is an open-source observability framework that provides a vendor-neutral standard for generating, collecting, and exporting the three pillars of telemetry data: logs, metrics, and traces. This standard eliminates vendor lock-in by allowing developers to instrument their code once and send the data to any compatible observability platform. Large observability vendors like Dynatrace and Splunk have been major drivers of the standard, actively contributing to and adopting it to ensure their platforms can ingest and utilize the unified OTel data. This massive industry and community support has propelled OTel to become one of the most active projects in the Cloud Native Computing Foundation (CNCF), often ranked second only to Kubernetes in contribution velocity, reflecting its rapid ascent as a fundamental open-source technology.
Against this open source backdrop, the mainframe has, for years, functioned as the IT estate's great anomaly, a high-value black box critical to global transaction processing yet largely opaque to modern operations teams. When a transaction hits a slowdown, the subsequent investigation often devolves into siloed, time-consuming "control room" investigations that span distributed and mainframe specialists. This is not sustainable. The announcement of IBM Z Observability Connect 7.1.0 is not merely a product update; it represents IBM’s formal capitulation to and adoption of the OpenTelemetry standard as the required interoperability layer for hybrid IT.
The market has been loudly demanding vendor-agnostic visibility. OpenTelemetry, or OTel, provides a standardized framework for collecting and managing the three pillars of observability—traces, metrics, and logs—removing the need for proprietary agents and conversion headaches. IBM’s embrace of this standard is a necessary move designed to maintain the relevance of Z systems within the broader cloud-native and DevOps movements. Without this integration, the mainframe risks becoming an operational liability rather than an engine of business resilience.
The core challenge has always been taking the richly detailed but highly proprietary performance data from z/OS—like SMF records and log data—and making it consumable by systems designed for distributed, cloud-native telemetry. The 7.1.0 release is architected to solve this exact problem. I view this as an essential, high-stakes move for IBM.
What was Announced
IBM Z Observability Connect 7.1.0 introduces several key architectural and functional changes. The solution aims to deliver comprehensive visibility by combining three core mechanisms. First, the ZAPM Trace Components, formerly known as Z APM Connect, are designed to collect trace data from vital z/OS subsystems, including IBM CICS Transaction Server, IBM IMS, IBM MQ, and IBM Db2. These traces are then routed through the ZAPM Distributed Gateway, which supports exporting spans compliant with OpenTelemetry Protocol (OTLP) for ingestion by OpenTelemetry Collectors or proprietary systems like IBM Instana and AppDynamics.
Second, the Telemetry Controller is introduced as a custom, IBM-developed OpenTelemetry Collector specifically engineered for z/OS environments. This controller is a pivotal feature. It includes a custom receiver designed to consume metrics and z/OS log data streamed from the Z Common Data Provider. The Telemetry Controller’s primary function is to transform this proprietary IBM Z data—SMF records and syslog information—into the standardized OTLP format for both metrics and logs. This process aims to streamline the data pipeline, ensuring that all three telemetry signals—traces, metrics, and logs—are delivered seamlessly and consistently to enterprise-wide observability platforms. The design provides flexible deployment options, suitable for both evaluation and large-scale enterprise production. The overall architectural goal is clear: provide high-fidelity mainframe telemetry that can be processed without requiring specialist mainframe tools for initial triage.
The significance of this release is found less in its novelty and more in its necessity, particularly when contextualized against the competition. This segment of the market is becoming fiercely competitive. Dynatrace has been building out native support for z/OS components like CICS and IMS, and they actively leverage OpenTelemetry to stream mainframe logs. Splunk is also making moves. Broadcom, through its WatchTower Platform, is also heavily invested in using OTel to achieve unified observability across hybrid environments. Similarly, BMC is enabling OpenTelemetry metric streaming from its AMI Datastream for Ops solution to multiple backends like Splunk and Datadog. This signals that OpenTelemetry has become the new table stake for mainframe vendors. It is the language operations teams require to truly operationalize the mainframe alongside the cloud. IBM is not pioneering the concept of mainframe observability but is instead catching up to the widespread vendor and customer adoption of the OTel standard. Their success will depend on the performance impact and the completeness of their Semantic Conventions for z/OS specific data.
IBM: Not the Only Game In Town
Broadcom entered the OTel arena last year with its WatchTower Platform that aims to serve as a critical bridge, modernizing z/OS and related environments by integrating them into the broader enterprise observability landscape. It intrinsically focuses on AIOps to address the complexity of the mainframe, utilizing ML against vast streams of SMF, RMF, log, and CICS/DB2 transactional data to deliver proactive correlation and deep root cause analysis within this high-fidelity environment. Central to its modern strategy, WatchTower acts as a sophisticated OpenTelemetry (OTel) Exporter and Collector for the mainframe, normalizing and streaming this traditional operational data as standard OTel traces, metrics, and logs. This OTel normalization is a paradigm shift, effectively making the System of Record (SoR) legible to off-platform SRE and DevOps tools, thereby eliminating the proprietary data formats that have historically been a barrier to End-to-End Service Assurance. Ultimately, the platform empowers teams to manage the mainframe not as a siloed resource, but as an integral, fully observable component of cloud-native and distributed business transactions.
BMC is also actively enhancing its mainframe observability capabilities by leveraging its AMI Datastream for Ops solution to enable direct streaming of OpenTelemetry (OTel) metrics. This strategic move involves converting proprietary mainframe performance and utilization data into the standardized OTel format, thereby allowing customers to easily send high-fidelity metric streams to multiple, diverse observability backends like Splunk, Datadog, and others simultaneously. By embracing the OTel standard, BMC effectively breaks down the traditional monitoring silo, providing operations teams with unified, real-time mainframe visibility directly within the distributed tracing and analysis tools they already use. This vendor-agnostic approach significantly improves AIOps initiatives by standardizing the data pipeline and integrating mainframe operational insights into enterprise-wide service assurance strategies.
Looking Ahead
The most critical element of this announcement by IBM and the wider adoption trend by the mainframe community is not the technology itself—as OTel is an industry standard—but the cultural and organizational impact it facilitates. This move aims to deliver a common vocabulary for distributed and mainframe teams, which is arguably more important than any line of code. The key trend that I am going to be looking out for is the progress of the OpenTelemetry Mainframe Special Interest Group. IBM has positioned itself as a key contributor here, but the standardization of z/OS data models (Semantic Conventions) for things like Job Processing and System Performance Metrics will ultimately determine how effective these tools are. If the community fails to align on consistent data definitions, the cross-platform value proposition diminishes rapidly.
My perspective is that IBM is navigating a tightrope walk. On one side, they own Instana, a full-featured observability platform that competes directly with the likes of Splunk, Dynatrace, Datadog, and many others. On the other, IBM must provide vendor-agnostic output to satisfy customers who already have substantial investments in other third-party backends. The Telemetry Controller’s ability to efficiently normalize and export metrics and logs without incurring excessive overhead is paramount. Going forward, I am going to be closely monitoring how the company performs on two fronts: the adoption rate of the Telemetry Controller outside of Instana ecosystems, and the development of language SDKs for COBOL and PL/I, which is where the real depth of mainframe application tracing resides. When you look at the market as a whole, the announcement validates the tectonic shift from closed, proprietary system monitoring to open, vendor-neutral observability, forcing every legacy mainframe software player to either conform or risk irrelevance.
Steven Dickens | CEO HyperFRAME Research
Regarded as a luminary at the intersection of technology and business transformation, Steven Dickens is the CEO and Principal Analyst at HyperFRAME Research.
Ranked consistently among the Top 10 Analysts by AR Insights and a contributor to Forbes, Steven's expert perspectives are sought after by tier one media outlets such as The Wall Street Journal and CNBC, and he is a regular on TV networks including the Schwab Network and Bloomberg.