Browse Prior Art Database

Fault Tolerant Content Index Processing in a Content Management System

IP.com Disclosure Number: IPCOM000247060D
Publication Date: 2016-Jul-31
Document File: 2 page(s) / 102K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a method to handle content index processing with fault tolerance for an enterprise content management system.

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

Page 01 of 2

Fault Tolerant Content Index Processing in a Content Management System

Enterprise content management systems are facing a common challenge in dealing

with growing data. Traditionally, the infrastructure of an enterprise content management system comprises a content index server for handling meta data of content and one or more object stores for content object storage. Distributed object stores and file systems are the underpinnings of content object storage. Scalability and efficiency of content object storage have already been the focus of much work. A

majority of systems are designed to leverage a cluster of commodity hardware. Each node in the cluster is a commodity server with many locally attached hard disks (i.e., a storage node). The cluster scales horizontally as more storage nodes are added.

A method is still needed, however, to efficiently handle content index (i.e., meta data) and achieve fault tolerance for the content index server. A content index server may

become a single point of failure (SPOF) if fault tolerance is not properly handled. The size of the challenge increases with the exponential growth of content data.

The novel contribution is a method to handle content index processing with fault tolerance for an enterprise content management system.

In the traditional approach, if the content index server fails, the entire enterprise management system becomes useless. When this happens, users cannot perform creation/update/retrieval/deletion of content due to inaccessible content indices.

The disclosed approach allows continued operations of creation/update/retrieval/deletion of content when a content index server fails for a configuration of three content index servers. To support a situation of two server failures, a configuration of minimal five content index servers is required.

A global consistency server is proposed to keep logging of the meta-data and keep tracking the status of the meta-data replications among all content index servers. This global consistency server is built on top of a fault-tolerant system that has no single point of failure (e.g., HBase). Data logging is achieved by storing a set of data records into the datastore built in the fault-tolerant system (e.g., HBase). The status tracking is a field/property of each data record. It utilizes synchronous replication (parallel

writes/updates) to the three identical content index servers (default of three) to achieve the removal of SPOF and achieve fault tolerance. It also uses relational databas...