by Dawn May | Apr 23, 2026 | AI/ML, Performance, Systems Management
i-Rays brings observability to IBM i
A buzzword in IT today is “observability”. As IBM describes it, “Observability is the ability to understand the internal state or condition of a complex system based solely on knowledge of its external outputs, specifically its telemetry.” I have a more simple description – it’s the ability to understand what is happening end-to-end in distributed business transactions. The technology surrounding IBM i has been focused on observability to understand the business workload. IBM i often sits in the middle, a sometimes overlooked centerpiece in the observability puzzle. The business data, the crown jewels of the business are, in Db2 for i.
There are various technologies in the world for gaining observability in your environment. Dynatrace is popular and many IBM i shops use Dynatrace to better understand the business transactions that surround the IBM i. However, observability of the i itself is limited. Dynatrace offers a default extension for IBM i, using various IBM i Services to access system information. While these services are easy to use, the extension itself can lead to performance problems, particularly when identifying top jobs consuming CPU. I’ve encountered several Dynatrace monitored systems suffering performance issues due to Dynatrace native extension for IBM i commanding the most resources. Generic platform vendors often deliver generic solutions, without understanding the intricacies of IBM i architecture. This is not a problem specific to Dynatrace alone — it is about a lack of familiarity with the unique architecture of IBM i, which is not common knowledge.
Today i-Rays offers an alternative Dynatrace extension. Built by IBM i performance experts who leverage the built-in performance data already available, you install nothing and enjoy low overhead. Couple excellent data collection capabilities with machine learning and proprietary AI, i-Rays can identify unexpected changes in performance signatures to stop problems before they appear.
If your IBM i shop has embraced Dynatrace, I recommend looking into i-Rays as an alternative to the default extension, to maximize observability of your most important asset.
For teams not using Dynatrace, iRays can operate as a standalone solution or integrate with any monitoring platform through OpenTelemetry.
by Dawn May | Apr 13, 2026 | AI/ML, Navigator for i, Db2 for i, Performance, Systems Management, Troubleshooting
IBM i sits at the core of many businesses, small and large. It quietly runs business transactions and manages the crown jewels of the company: business data. IBM i is surrounded by other technologies accessing that data without a thought. It quietly runs the business, often tucked away with little attention paid to it. Just one or two administrators versus teams of people managing everything around it. IBM i simply works. Until it doesn’t.
Ease of use is often cited as a benefit of running IBM i. The tightly integrated nature of the system makes it easy to manage. In one box you’ll find the base operating system, advanced networking, HTTP servers, file systems, and critically, the database. It simply works, but under the covers, the system is quite complex featuring hundreds of parameters for tuning and controlling the system, with shipped defaults that generally work well, but not always. The learning curve required to manage the system and keep it running well can be steep. Very often tuning is done with a “set it and forget it” mentally, but workloads shift and change. Minor adjustments accrue over time making tuning an ongoing necessity.
Understanding all the parameters, identifying which ones are important and which ones need to be adjusted can be daunting for those without years of experience. Even experienced administrators are often unaware of key tuning features. There’s just so much to know that one person simply cannot know everything.
The Value of IBM i System Data and System Configuration
IBM i collects incredible amounts of system data to assist tuning decisions. You may not even be aware of this data. Collection Services, the SQL Plan Cache, Index Advice, and optional Job Watcher data all provide insight into how the system is running. But understanding this data is also daunting. What do charts of Collection Services data presented in the Investigate Data task of Navigator tell you? With hundreds of advised indexes, which ones are the most important to create?
Yes, everyone should increase their knowledge of the platform they are working on, but I also encourage leveraging tools to make your job easier. Software advances in AI and ML can and should be used to help. New software that can understand IBM i like never before is available. i-Rays is built to understand and exploit IBM i performance data including Collection Services, the plan cache, and more. With built-in AI/ML, i-Rays learns your typical performance signature and then identifies anomalies. In addition, i-Rays can help identify what needs to be tuned to improve system performance. Over time, i-Rays can automatically tune the system for you.
by Dawn May | Apr 18, 2018 | ACS, Performance
This week, I want to review considerations to copy performance data to a different partition. Many IBM i shops may want to analyze performance data on a partition other than the production system. The Performance Data Investigator uses SQL, and it may be desirable to keep that SQL workload off your production partition.
(more…)
by Dawn May | Feb 21, 2018 | Subsystems, Performance, Work Management
I’ve written about workload groups in the past in Workload Group Configuration with IBM i 7.3 and Manage Workloads better with Workload Groups.
This blog reviews how to understand possible performance implications when using workload groups.
(more…)
by Dawn May | Aug 1, 2017 | Subsystems, Performance, Work Management
The End Subsystem (ENDSBS), End System (ENDSYS), and Power Down System (PWRDWNSYS) commands all have the End Subsystem Options (ENSSBSOPT) parameter. This is nothing new—it’s been there for a long time. In case you are now aware of this parameter, this blog is for you.
(more…)
by Dawn May | May 22, 2017 | Navigator for i, Performance
With the 7.3 release, IBM introduced historical performance data collection and the ability to view your historical data graphically with Graph History. With the recent Navigator service pack (HTTP Server Group PTF SF99722 Level 7 or later), graph history was expanded with stacked charts. Stacked charts allow you to display and compare a historical data metric over days, weeks, months, or years.
(more…)
