InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Method and System for Providing Queue Monitoring Tools for Advanced Capture of Out of Memory (OOM) Issues

IP.com Disclosure Number: IPCOM000240858D
Publication Date: 2015-Mar-06
Document File: 2 page(s) / 104K

Publishing Venue

The IP.com Prior Art Database


A method and system is disclosed for providing queue monitoring tools for advanced capture of out of memory (OOM) issues.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 01 of 2

Method and System for Providing Queue Monitoring Tools for Advanced Capture of Out of Memory (OOM) Issues

Gathering data for out of memory (OOM) issues in Business-to-Business (B2B) integrations, a snapshot of the state of a computing system leading up to the OOM is very useful. As the OOM itself is destructive, a snapshot after the issue occurs is of less useful. This makes tracking down intermittent issues difficult. Existing solutions are able to gather the snapshot on a regular basis in hopes of catching the issue. The drawback there is the snapshots take up a lot of space. Snapshots collected with a short interval between collections result in issues related to memory and storage. Choosing a longer interval is less likely to capture the issue as it is happening. When the OOM is not predictable this makes it hard to gather the data and often results in many system outages before getting the data needed to address the issue. The issues often repeat quite early and there is need for a method and system for capturing snapshots a short while before an OOM exists.

Disclosed is a method and system for providing queue monitoring tools for advanced capture of OOM issues. The method and system utilizes a monitoring tool that looks at specific indicators that predict upcoming OOM issues and starts gathering related data before the issue happens. Presence of the specific indicators initiates monitoring tools to gather snapshots of diagnostic information when OOM issues are likely to happen thereby reducing the amount of space taken up. Specific indicators that primary indicate an OOM condition are queue depth indicator and time on thread indicator.

In accordance with the method and system, while a B2B integrator is running in a normal condition, the queue depth for the work queues stay within a predictable range. Further, an OOM condition is identified by one or more queues climbing to many times its usual size. Subsequently a workflow engine in B2B Integrator takes workflows off the queue and executes them. The time on thread stays within a predictable range and an OOM is often proceeded by a workflow staying (stuck) on thread much longer than usual which usually indicates OOM issues. The threads that are "stuck" and the workflows waiting on queue are not always the actual issues but are often victims of the issues that lead to OOM. However, getting a snapshot of the state of the...