• Exceptions: provides a complete vision of the most important hardware and software threshold violations which helps locate problems and anomalies immediately. The default thresholds fit well for most installations.
• Configuration: provides a detailed vision of the hardware and software configuration including virtual machine configuration; this information is the first mandatory step in order to control and manage your systems and in general to perform all Capacity Management activities. EPV correlates data coming from various sources and systems, producing a global vision of shared resources such as CECs, in a completely automated process. EPV audits configuration changes allowing an immediate identification of possible anomalies related to them.
• Resources: provides a complete vision of the “health” condition of the critical hardware resources, especially those shared amongst different z/VM systems as processors, memory, disks and channels. Starting from an integrated enterprise view you can navigate and analyse any saturation or performance problem. The time period and resource causing the bottleneck are highlighted using a red background so they can be very easily located.
• Workloads provides a detailed vision of your workloads, allowing a simple and guided analysis of what is going on in complex environments through drill-down capabilities. Starting from a FUNCTION view, performance and resource consumptions can be analysed drilling down to virtual machines details. Information about the top processes running inside Linux guests is also provided.
• Trends provides daily, weekly and monthly views of performance and resource consumptions at the enterprise, system and virtual machine levels. By use of these views it is possible to understand the growth of your workloads and its impact on your systems.
Access to available data
EPV for zLINUX uses standard measures normally available in any z/VM environment and supports various sources, such as ITRM (IT Resource Management), MXG (Merrill’s eXpanded Guide) or native z/VM monitor records. EPV optimises data collection by using existing summarised data, limiting the number of variables, using intelligent filters and avoiding data duplication and redundancy. The product’s architecture is modular and user friendly. EPV can be installed on most of the popular hardware and software platforms available on the market.
Using “de facto” standard technologies
All reports are produced in simple static HTML pages published on a server of your choice and can be accessed by your favourite “browser”. All the tables provided in the reports can be exported individually or together in a MS-Excel spread sheet.
EPV for zLINUX architecture is based on the following three layers:
• An interface to collect input data;
• An engine to correlate and aggregate data;
• An engine to produce HTML pages.
The interface to collect input data is designed to optimize performance and consumptions of the loading phase, avoiding duplication or loss of data. It uses simple exits that adapt the product to the environment characteristics; a set of ready exits for each data type to acquire is provided.
The engine to correlate and aggregate data builds a standard SAS or SQL relational Data Base, containing only those metrics used by the product and automatically aggregated at the monthly, daily and hourly level. You can customize the product by setting the appropriate parameters and values to keep in your EPV Performance Data Base the desired time periods.
The engine to produce the HTML pages can be customize to:
• Report daily resource and workload statistics for any desired period
• Report daily, weekly and monthly trends for any desired period
• and much more…
EPV for zLINUX is available in two modes: standard or Plus.
The requirements for EPV for zLINUX are:
• Availability of a SAS System Base module at least on a PC
• Availability of one of the following products: SAS ITRM, MXG.
EPV for zLINUX Plus has no pre-requisites; it’s an “out of the box” solution which integrates a light version of the EPV zParser product to interpret the needed z/VM monitor records and all the other input data required.