i-Rays Observability and Tuning is a tool for analysing, evaluating and optimising the performance of the IBM i (aka AS/400) platform. It is based on a knowledge base of optimal system parameters (recommended settings) and on the detection of anomalies, understood as deviations from identified patterns of behaviour of a given system. A set of corrective parameters generated in real time allows the supervised system to be kept within the optimum performance corridor.
Learn moreIBM i (AS/400) Observability
i-Rays for the IBM POWER/IBM i platform is an aggregated behavioural analysis solution. It is based on statistical methods and AI/ML algorithms, being capable of capturing low-level dependencies between phenomena occurring in the system. It allow for not only standard monitoring, identification of sub-optimal system settings or job runtime parameters, but above all detects deviations from behaviour considered as typical in a given system - all this together with the generation of information about the detected problems/deviations and with reconfiguration hints/suggestions to maintain the optimal state of the software stack running on a given system (logical partition).
Key benefits
Proactive prevention of problems that may arise in the system (resource saturation, performance degradation).
Higher performance - by increasing the transactional throughput of the system.
Cost reduction - by being able to reduce the resources and system licences used.
Main features
Digital worker
It continuously supervises the system, detecting behavioral anomalies in the context of the business characteristics of the load. By filtering out possible false positive alarms, it frees staff from manual interpretation of results. The system is diagnosed simply as 'working normally' (regardless of the load) or 'requiring attention'.
Advanced advisor
Interpreting real-time metrics against the background of system historical data and the knowledge base, it identifies deviations from both the "general norm" of the business workload and from the "typical behavior of a given system". The advisor module also generates performance optimization recommendations - both for iOS and DB2, in the form of ready-to-use system commands.
Fixing
Recommended values (best practices and guidelines) are shown as references together with key metrics. The values are defined contextually, for the combination of the parameters of a given system and its software stack. General IBM POWER hardware metrics from HMC and VIOS are combined with data obtained from IBM logical partitions. This module also allows for the analysis of load trends to support investment decisions - e.g. system sizing.
Main use cases
Detection of non-standard behaviour
Broken down into abnormal load (including business load) and abnormal system response to standard business load (Anomaly Detection module).
Real-time generation of optimisation tips
in the areas of both the operating system with its software stack, and the database (System and DB Advisor module).
System monitoring
Presentation of selected performance metrics - against a set of guidelines and recommendations based on system characteristics (Smart Monitoring module).
System optimisation
Especially if there is no possibility to upgrade the system - because it is already at its maximum available configuration or upgrade is too costly.