Browse Prior Art Database

Using Empirical Data for Managing Applications in a Cloud Environment

IP.com Disclosure Number: IPCOM000246419D
Publication Date: 2016-Jun-06
Document File: 5 page(s) / 191K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a method and system to leverage data and analytics gathered across all tenants and applications in a cloud to improve the management of specific applications of specific users. This is a community knowledge approach.

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

Page 01 of 5

Using Empirical Data for Managing Applications in a Cloud Environment

In public cloud environments, systems and administrators must manage thousands of applications belonging to different tenants or users. These thousands of applications contain clusters of applications that are very similar (i.e., have a similar topology, similar configuration, use similar software components, etc.). Gathering data (e.g., monitoring, event recording, change recording, etc.) and then performing analytics on that data provides single users of single applications insight into the application's performance. That insight can be applied to improve the applications. However, this data is kept private per tenant/user and application. Knowledge is not leveraged across the whole cloud to help other users improve associated applications.

Figure 1: Cloud Hosting Environment - Topology Clusters

The novel contribution is a method and system to leverage data and analytics gathered across all tenants and applications in a cloud to improve the management of specific applications of specific users. This is a community knowledge approach. The proposed system and method use empirical analytical data obtained across all applications in a public cloud to improve the management of specific applications.

In a Learning phase, the system:

1. Collects data from all applications: topology data, events, performance metrics, changes

2. Uses analytics to detect performance anomalies (i.e., bursts or drops), including remediation points

3. In case of detected anomalies and the remediation of said anomalies, captures snapshot of a time window of all application data and stores it in a global knowledge base

In an Execution phase, the system:

1. Collects application data 2. Uses analytics to detect performance anomalies 3. In case of an anomaly, looks up entries in the knowledge base with the same kind of anomaly and the same kind of application topology, and then applies the

1


Page 02 of 5

  recorded remediation to current application 4. In case of pending changes, looks up entries in the database with the same kind of change, same kind of application topology, and a recorded resulting anomaly to advise on the expected effect of change

Figure 2: System Overview

Figure 3: Service Topologies - Data gathering

Detailed Flow of the Invention - Learning Phase (continuous)

1. For each application of each tenant, collect application instance data: A. Topology data: components (e.g., platform included, packages and all versions), relationships, component configuration

B. Topology changes (e.g., configuration change, component version upgrade, add/remove components)

C. Events (e.g., threshold breaches, alerts, connection events, etc.)

D. Health, resource usage and performance metrics (e.g. system load, response times, transaction times, etc.)

2. Health, resource usage, and performance metrics (e.g., sys...