Browse Prior Art Database

An impact analysis method of system patches and customer applications

IP.com Disclosure Number: IPCOM000246404D
Publication Date: 2016-Jun-06
Document File: 4 page(s) / 120K

Publishing Venue

The IP.com Prior Art Database

Abstract

In system software and application software system lifecycle, the maintainence patches will often be applied to provide bug fixes and function enhancements; but how to evaluate the impaction of new patches is awlays a problem. Here , we provide a new method to evlulate the impaction by analyzing system runtime data, it gives the risk report on both functional risks and performance risks based on test results .

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 4

An impact analysis method of system patches and customer applications

In large IT system, there is a big challenge to maintain system stability with a lot of system patches, usually IT service professionals provide service to apply system fix packs. Before fix packs are applied, enterprise clients concern about the impact of fix pack on their applications, especially impact on performance. (If fix pack is not applied, known issues are left in system. But fix packs may bring new issues or affect system performance. )

So enterprise clients ask IT service experts to provide risk evaluation report to estimate the impact of fix packs. The risk does not only mean functions issue, and also means performance downgrading. Enterprise clients care about performance regression, and will have concern if performance is 10% downgrading. And performance issues are more difficult to find out in lab.

This is really tough work to service experts. A large IT system usually includes hundreds of fix packs in each operation system release. Service professionals have to analyze hundreds of fix packs, provide the risk evaluation report of fix packs and decide which fix packs to apply in customer production system.

One example of working process is like this:
1. Prepare a group of latest patches.

2. Review patches by expert
3. Apply patches in test environment
4. Execute test cases for one month
5. Evaluate both functional risks and performance risks based on test result
6. Remove high risk Patches
7. Apply on production system

There are two kinds of difficulties:
1. How to choose a suite of test cases to verify fix packs. Expert has to decide the key functions to test by experience.

2. How to find high risk patches. It is difficult to find high risk patches. Nobody knows which patches cause issues/performance downgrading. Experts have to choose a patch by experience and apply/remove patch to compare the result. This is really tough and not efficient.

We consulted service experts in one big bank customer site. There are no any tools to provide runtime material to support risk report. Service expert has to depend on person experience to evaluate fix packs impaction and decide the verification scenarios.

Our method is going to provide an impact analysis method that using system runtime sampling data to assist risk evaluation.

The performance weight can be calculated from sampling data, for example when sampling on CPU, we can get we calculate the number of times that s...