In today’s IT infrastructures the rate of change and business demands require that the IT team resolves problems quickly and accurately--while preserving SLAs. Some problems are difficult to detect where it may build up over time, while others are difficult to diagnose as the amount of data may be too large to consume manually in a timely manner.
IBM zAware provides cutting edge pattern recognition technology that provides a view of system behavior in near real time, across all servers in the enterprise. It analyzes massive amounts of data in the message logs from z/OS and Linux on z Systems (with z13 servers)1 quickly. IBM zAware uses machine learning with historical data to proactively detect anomalies when systems deviate from normal behavior. IBM zAware can help identify unusual behavior quickly, reduce the mean time to recovery, at the same time reducing the time and skill to detect the problem. It can help diagnose major problems while they occur or heighten awareness of smaller problems before they become a bigger problem. While traditional automation solutions help detect 'what you know', the big strength of IBM zAware is that it can help detect 'what you don't know', whether it is during normal operation or after change or service window. If the anomalous behavior turns out to be a problem, you can add the resolution to your process/procedures to avoid the problem in the future.
IBM zAware learns about 'normal' behavior from historical data, to create a model. Then, in real time, it analyzes and scores each message against that model, to identify abnormal or anomalous behavior. The message may be IBM or non IBM, or a user application message. The scores can be visualized via heat maps and detail views in a browser based GUI. IBM zAware can detect if a message is new, or different because of the time it occurred, or if the volume or pattern of messages has changed, or if the context in which a message normally occurs has changed. For each monitored system, scores are generated for the messages as well as for time intervals. To feed into the ecosystem, APIs are available for scores to be consumed by higher level system management products, IBM or non-IBM.
According to several clients IBM zAware is easy to install, activate and get the analytics going. A customer stated that it has helped maintain SLAs by helping detect and hence resolving the problem quickly, which in some cases may have resulted in monetary fines for missed SLAs. In another case it identified a problem after a change window which was resolved very early and quickly, that, if left undetected would have caused bigger problems and impact to business later. One customer even saw a better synergy in solving problems across multiple teams, as it identified the area with abnormal behavior quickly and saved diagnostic time across multiple teams.
This was just a quick introduction on how IBM zAware can help with its anomaly detection capability. Please feel free to reach out for a detailed discussion and help. Read more about IBM zAware capabilities at https://ibm.biz/BdXnrv