Having to respond to false alarms can be very frustrating for mainframe IT. Your staff may not have the time or the expertise to set new thresholds when they’re needed. The thresholds that have previously been set up – perhaps many years ago — may be outdated and inaccurate. This can cause IT to waste time and effort in tracking down alerts only to discover that there’s not even an issue.
Or worse yet, inaccurate or often static thresholds are set very high to accommodate a variety of possibilities based on certain days of the week, special processing requirements, and other situations. This means a problem could be ongoing for hours before you are alerted that something is seriously wrong. Your business could suffer because resources aren’t in place to support critical services. Having better analytics can help solve these challenges and reduce guesswork.
1. Dynamic workloads require dynamic thresholds
In the past, IT often defined thresholds based on known peaks and that worked out well. But today’s digital workloads are dynamic, volatile, and driven by mobile activity. Examples include people looking at their bank balances, checking on cheap air fares, reading restaurant reviews, and comparing hotel prices and amenities multiple times a day, on weekends, and often late at night. Banks have expanded their view of peak times to include early weekend hours. That means static thresholds that are set to accommodate known peaks are no longer valid and traditional business cycles no longer exist.
Thus, a real issue that occurs at a much lower threshold goes undetected until hours later, when the static threshold number is finally breached. That approach can be costly. Alarms that are unable to distinguish between one-time anomalies and actual persistent problems provide a lot of false alarms, while true issues are lost in a sea of red lights.
Fortunately, MainView provides analytics to automatically create accurate dynamic thresholds based on your actual workloads and build calendars. MainView collects performance data on a daily, weekly, or monthly basis. It automatically does an analysis on this data and required calendars to create recommendations of dynamic thresholds to set based on our business cycles. This enables you to keep your thresholds current and reduce false alarms. Here is an example:
- Using MainView for CICS, you select a CICS region to identify which metrics and data to use to determine the most effective threshold settings. MainView does an analysis, provides threshold recommendations, and gives you the ability to automatically set these recommendations. This capability is all part of a closed-loop process that delivers intelligent, proactive automation to provide systems management that can meet the challenges of an always-on digital world.
2. Defining alarms can be done easily
You can define your own alarms but BMC also provides out-of-the-box smart alarms for mainframe databases and systems based on best practices, which simplifies the process. There are a couple of different approaches to setting alarms and both are valid. One casts a wide net and defines alarms for all potential problems, whether or not they have actually happened. That’s a proactive approach that covers all the bases. Another approach reflects problems that have been experienced and analyzed and then alarms are defined to prevent them from occurring again. This is an additive process, so each time an issue occurs you will need to create an alarm to handle it.
3. Move from observation to action
Staying competitive in digital business requires evolving from observing alarms to addressing exceptions and automating them. MainView triggers automation on any detected exception throughout the MainView solution set. When an exception is detected, automation takes an action. Dynamic thresholds based on business cycles provide accurate exception detection and meaningful monitoring in off-peak as well as peak cycles. This reduces the guesswork and time required for establishing and maintaining thresholds.
Don’t let false alarms and unexpected business cycles impact your organization’s ability to compete in the digital economy. To learn more about how to maximize the value of mainframe automation and analytics, see this customizable demo.
- Data Management Best Practices: 6 Essentials for Your Mainframe Team
- Sharing Best Practice: IMS Availability During Change
- 3 Wicked MLC Moves to Reduce Your Mainframe Costs
- Where’s the Ops in Dev Ops?
- It’s 2016: Is your Mainframe Ready for the Digital Era?