Use ObjectStore's Cache-Forward Architecture to Get Data Where it Needs to Be

Progress® ObjectStore® has a Cache-Forward™ Architecture (CFA) that is a distributed set of components that delivers enterprise data to performance-critical services fast - and in a format tuned to their requirements. Designed to provide optimal balance between core memory and secondary storage, the CFA leverages the existing database of record to continuously move data out to the edge using an object model optimized for a particular use case.

Figure 1: Example of the Cache-Forward Architecture

The Benefits of the Cache-Forward Architecture

Regardless of your deployment - an SOA-based web service, high traffic website, or cached data layer in a distributed grid - the CFA affords high performance, linear scalability and reliable access to critical data. Unlike simple caching solutions, the CFA continuously and seamlessly moves data from multiple sources to performance-critical applications - directly to memory at the edge, in the appropriate object model. The combination of moving data to critical services with transformation to the appropriate object model offers unmatched performance and agility while insulating mission critical databases of record from the impact of change and high demand.

Provide Services with Data at In-memory Speed

Localize Data: Distribute your data with your process. Deploy a distributed cache with a service and ensure that data is available where it's used. The cache is maintained in memory, reducing, even eliminating, network and disk access and significantly lowering latency. Deploy multiple instances of a service with specific use cases and smaller footprints for increased throughput. Network and disk access are your biggest bottlenecks - don't let them throttle your services.

In the Optimal Format: Caching a relational result set isn't enough. Bridging the gap between a complex object model and a relational model requires additional time and effort. In most cases, a complex model requires a large number of joins across tables or may even be impossible to model - both of these characteristics prevent optimal performance and increase maintenance costs. An object-oriented cache ensures that persistent data structures and relationships directly align with the use case of the service offering.

Figure 2: Align your data with your performance requirements. Map your relational data to the optimal model for a given use case.

Get information to the right place

Continuously Up-to-date: Accurate decisions need accurate data; if your cache has stale information it's worthless. ObjectStore's CFA keeps the data in your cache current with low latency updates from multiple heterogeneous data sources. Your employees, customers and services have the most up-to-date information for every transaction.

Figure 3: Automatically, continuously and with low impact keep your cache up-to-date.

In a complete manner: As business grows so do the quantity, type and locations of data necessary to run it. The CFA integrates multiple heterogeneous data sources to supply the content of a cache - regardless of the format or source. Your cache will contain a coherent, unified and complete view of the data you need to make decisions.

With Agility: Gain the flexibility to easily maintain existing cache and data source integrations as well as quickly offer innovative new services with new sources. The effort to maintain consistency of multiple heterogeneous sources throughout the lifecycle of your services can be daunting. The effort required to maintain more than a few simple point-to-point integrations will quickly become a black hole for your time and resources hindering the ability to respond to changing business requirements.

The CFA allows you to manage and understand dependencies as your business requirements change. With our innovative exchange model you can avoid point-to-point dependencies and actually gain efficiencies as your integration needs grow. The cache dramatically improves the response time of your data. Feature-rich, eclipse-enabled tools allow you to reap the same benefits for both your development and maintenance efforts.

Ensure the Availability of Your Data

Figure 4: The CFA integrates multiple heterogeneous data sources to supply the content of a cache - regardless of the format or source.

Meet your customer service requirements - You use a cache because your business critical applications require up-to-date and complete information now. For these applications your cache must be highly available and offer seamless failover. A durable cache allows the state of one process to be transferred effortlessly to another with no loss of state. Make sure you're always there to meet the demands of your customers.

With available resources - Efficient, scalable deployment of many services requires a smaller memory footprint per service or more hardware. Adding hardware is likely to be neither cost effective nor operationally feasible. A durable cache automatically moves the appropriate data from disk into memory and back out again providing a smaller footprint to leverage existing hardware.

Get More Value from Your Existing Database of Record

Let the database of record be your database of record - The database of record should maintain the canonical view of your data - suitable for business intelligence as well as low impact transactional operations. It should be optimized and structured in the manner that best meets those goals. Don't waste time and money scaling up your relational database in an attempt to meet high performance, transactionally-intensive demands. ObjectStore's Cache-Forward Architecture insulates your database of record from frequent change and high demand by offloading the work to a cache designed to meet your goals, leaving you with a more agile, maintainable development and deployment process that will exceed your performance requirements with less effort.

Related Topics

• ObjectStore Enterprise
• Object Data Management
• Data Integration Software
• SOA Infrastructure