IBM Z and LinuxONE - Software - Group home

Announcing IBM Z Performance and Capacity Analytics

By Chris Walker posted Wed March 04, 2020 05:05 PM


On February 18, 2020 IBM announced IBM Z Performance and Capacity Analytics V3.1, an complete solution for performance analysts and capacity planners  to track key performance metrics and make informed decisions about future workload and Z environment changes.

The new release builds upon significant updates delivered throughout 2019 including new reporting based on Cognos Analytics and new capabilities that support Tailored Fit Pricing strategies such as managing MSU usage against the defined baseline, forecasting future consumption and identify optimization opportunities.

In this article, we’ll share a brief overview of some of the new features delivered in the new release:

Addressing the question of “What If” in capacity planning

For capacity planners, a major part of their responsibilities is to provide informed judgement on the future needs of the mainframe environment to support current and future workloads. This may include providing guidance on future hardware upgrades and other capital expenses to support business goals. This is aim around the new “What If” capabilities built into IBM Z Performance and Capacity Analytics. Current performance can be measured accurately with the existing configuration but what benefit might be we see if we were to change this environment or make a decision to upgrade our processors to a newer model – can we forecast the potential benefit before making the investment decision?

Cognos Analytics report showing the What If capabilities within IBM Z Performance and Capacity Analytics V3.1

A brand new Cognos dashboard shows how CEC and LPAR capacity varies when:

(a) the number of physical processors is increased or

(b) a processor model is upgraded.

From here, we can visualize in an easy to digest fashion the effect of upgrading to new processors by showing how much additional capacity would be available on an overall system level as well as an individual LPAR level.

Lowering the cost with additional zIIP-eligible capabilities

Processing SMF data in near real-time is essential to get timely insight into data. In previous releases, the automatic data collectors that move the data from the spoke LPAR systems to the hub and the forecasting modules were defined as zIIP-eligible. In the new release, the continuous collector process has also been updated so that it now can be executed on zIIP engines if available. Internal testing has indicated this can save up to 70% of the processing by the continuous collector will consume less MSUs when curating data, which is especially applicable under enterprise consumption Tailored Fit Pricing model through reduced resource consumption.

“Smart Path” LPAR MIPS Usage Profiling

As a performance analyst, tracking performance over an extended period of time is critical for understanding trends and establishing if there is a change in behavior, either in the short-term due to spiky or unpredictable workloads, or a long-term change in performance.

The new Smart Path report builds an LPAR-level profiles of expected hourly behavior based on historic usage, providing insight to LPAR hourly MIPS usage, for every day of the week.

A Cognos Analytics report showing SmartPath capabilities in IBM Z Performance and Capacity Analytics

On the example shown here, the calculated profile together with an upper and lower bound for expected MIPS usage each hour is plotted as a line chart. The actual peak MIPS usage recorded within each hour period is marked against this as a bar making it clear and obvious when performance does not align with expected behavior.

Enhanced exception reporting

Exception reporting enables tracking of unusual conditions that are discovered when the input data is being analyzed. Using defined thresholds tested against the Key Performance Metrics, if any are breached then a record is noted on an exception table that can be reviewed by the performance analyst. New updates make this feature more accessible and practical. Now if an exception occurs, a notification WTO message can be written to the console indicating which threshold was triggered, what is the defined value of the threshold and the actual value found in the processed SMF record. The console message can be leveraged by other tooling, for example automation, if something should be triggered when an exception occurs.

Additionally, a new report has been created within Cognos Analytics that provides a summary of the exception analysis. By selecting a time period, you can see a breakdown of what exceptions have been identified, which systems the exceptions occurred on together with a trend analysis showing a day by day view that could indicate a long-term problem that needs addressing.

A Cognos Analytics report showing Exception details within IBM Z Performance and Capacity Analytics V3.1

From this summary you can automatically drill-down to more detailed reporting about the specific exception, when it occurred and how often we might have seen this problem in the past.

Improved reporting within Cognos Analytics
In 2019, reporting based on Cognos Analytics was introduced to give a modern, extensible interface to empower end users to access information quickly and efficiently with the additional ability to customize to their needs (A limited use license of Cognos Analytics V11.0.13 is provided with the product).

In this new release, there are now over 200 predefined reports and extensive work has been done to structure the reports logically within Cognos to make it easier to access summary details faster and then drill-down to details needed to analyze the root cause of performance issues or make informed decisions for planning.
Cognos Analytics report showing KPM Summary in IBM Z Performance and Capacity Analytics V3.1

Another example of improved reporting is with a new summary report pulling all the Key Performance Metrics reports into a single dashboard delivering a clear view of the performance of your z/OS environment across a selected time period.

Key Takeaways
IBM Z Performance and Capacity Analytics is an integral part of IBM’s portfolio for leveraging analytics to support the whole IT operations team. It is a core part of the IBM Z Operations Insight Suite, which additionally brings in IBM Z Operations Analytics for problem identification and anomaly detection using machine learning for operations. You can learn about the latest updates for both offerings in the Announcement Letter. Please reach out to your IBM representative if you have any questions or would like a demonstration.