There will be a system update on Tuesday, January 16th, 8 PM ET. You may experience a brief service interruption.
Browse Prior Art Database

Content and filename based application discovery

IP.com Disclosure Number: IPCOM000234991D
Publication Date: 2014-Feb-21
Document File: 4 page(s) / 54K

Publishing Venue

The IP.com Prior Art Database


Disclosed is a system for content and filename based application discovery that enables a user to identify the file listing associated with a given software/application.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 38% of the total text.

Page 01 of 4

Content and filename based application discovery

In discovery and migration, a method is needed that enables a user to identify the file listing associated with a given software/application. This information can be used to create the signatures of the software/application as well as provide information for migration failure diagnosis.

No current solutions address this need.

The novel contribution is a system for content and filename based application discovery. The components and process for implementing the solution include :

1. Using an application discovery tool to identify software/application installed at each image

2. Calculating equivalent classes for these software listings

3. Using set manipulation (i.e. intersection of images contain application x - union of images does not have application x) to get file listings associated with software bundle x

4. Using an alignment algorithm to align the full file name with the application bundle

5. If a discovery tool is not available, then the system can cluster file names to obtain a product bundle and continue steps 3 and 4

6. For product updates and patches, the system can use either content clustering or file name clustering to find the product bundle and continue steps 3 and 4

Example Embodiment #1: Application name determined by discovery tools

If the application name can be determined using an existing discovery tool , then applications/middleware installed at each server are first detected using an existing product discovery agent. Most of the existing product discovery tool use some sort of product signature (content) for product discovery and the results can be far from perfect.

The application bundles across the entire servers are calculated . The application bundle can be defined as those applications always installed at a server as a bundle . For example, for a given set of applications, X1, X2, 1 Xk, if any are installed at a server, then all of the applications from this set are installed at the same server . Thus, these applications are always installed as a bundle. The application bundle can also be referred to as "equivalent class". The bundle can be calculated, for example, using a bit5vector hash algorithm.

Each application bundle is the basic unit. The file name associated with this bundle can be derived by:

Given: Ai = files for imagei installed application bundle x

Bj = files for imagej do have application bundle x installed


Page 02 of 4

Thus, the (intersection of A1 ... Ai) minus (union of B1 1 Bj) is the superset of those files associated application bundle x. The equation can also be written as:

When there are sufficient images , fnoise can be ignored or handled by any noise reduction approach.

Based on the above approach, the base unit, x, is the software bundle. If the software bundle x contains three applications (e.g., A,B, and C) the user needs to identify which files belong to A, which files belong to B, and which belong to C. To dissect the appli...