Browse Prior Art Database

System Health indicators

IP.com Disclosure Number: IPCOM000030135D
Original Publication Date: 2004-Jul-29
Included in the Prior Art Database: 2004-Jul-29
Document File: 4 page(s) / 75K

Publishing Venue

IBM

Abstract

System Health Indicators The disclosed idea provides a method to build a unique index that can immediately account for the status of the whole IT system; the index is based on a set of sub-indexes, used to provide the unique detail for each defined "sub-system" component.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 55% of the total text.

Page 1 of 4

System Health indicators

A system enterprise can be described as composed by all its subsystems and applications like z/OS, CICS, IMS, DB2, Lotus Domino, Network SNA/IP, FTP, MQSERIES, RACF, DFSMS, etc (*)

For example can be considered a Customer having in his simple data processing site IMS, DB2 and MVS (z/OS) only. When the question: "How things went in the system?" is rised, then a check on an historical performance database repository is performed: on IMS data tables, then on DB2, then on MVS, then on any other involved subcomponent. There will be an idea, based on certain parmeters (system metrics) but not a certain and unique number for the whole system. This will not help to quickly understand how much well things are going in general, and in particular for the specific system components and how far from the optimum. Moreover it cannot be directly compared whit the situation of the last week or of the last month, for example.

The disclosed methodology provides a way to say for example: "The system is performing 40% of its total potential, 60% away from best; and this is due to IMS and DB2 performing only 60%, and 20% respectively. These bad performances in DB2 are due to...."
To obtain this result, first of all, the system components have to be identified. Then each component should be weighted according to its importance inside the system and customer core business. For example let us consider MVS (z/OS), DB2 and IMS weight in the order: 50%, 30%, 20% in the overall system picture (50+30+20=100).

How can be evaluated that MVS is acting well: the CPU consumption is less than 80% for ex. (Other indicators could be the real and virtual memory usage, the paging usage,and so on)

CPU= 80 % means MVS=100% system OK CPU= 90% means MVS=50%

CPU= 100% means MVS=0% system KO

For example IMS is acting well if the response time is less or equal 1 sec and bad if equal to th...