Dismiss
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

A consolidated cache data invalidation approach for distributed systems

IP.com Disclosure Number: IPCOM000202067D
Publication Date: 2010-Dec-03
Document File: 3 page(s) / 102K

Publishing Venue

The IP.com Prior Art Database

Abstract

This article elaborates on a system for a consolidated cache data invalidation approach for distributed systems to reduce the inter-node communication.

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

Page 01 of 3

A consolidated cache data invalidation approach for distributed systems

In-memory cache are commonly used as fast read-access caches in state-of-the-art web based systems to reduce computational and backend access latencies during request processing. This approach improves response times and throughput while minimizing data storage access. As an extension to this approach, clustered environments that serve data for read/write purposes utilize inter-node communication messages to keep in-memory caches up-to-date upon data modifications.

With growing application and data complexity, the functional requirement of having cached only authentic and up-to-date objects in memory requires advanced caching strategies to limit system resource utilization of in-memory caches. Cache-to-cache communication across node boundaries for data integrity purposes requires a significant amount network and processing resources during peak load times that cause application data modifications.

(This page contains 00 pictures or other non-text object)

Fig.1 - State-of-the-art invalidation processing in distributed environments

Figure 1 depicts such a complex application data caching setup that requires inter-node communication to maintain data integrity of cached objects.

Typically, data from different application components is cached in different cache instance. This data has often a semantic or dependency relationship that requires the eviction and update of a large amount of cached data objects for integrity.

As an improvement to this approach, an 'Event Distributor' can be used as message broker for invalidation events that encapsulate resource data.

While inter-node communication for the majority of caches is disabled, the Event Distributor colle...