Version 4.0 Adds Breakthrough Performance

With a newly enhanced communications architecture, version 4.0 now provides even faster performance for ScaleOut StateServer's production-proven distributed caching. This version also adds exciting new features that further enhance ScaleOut StateServer's APIs and management capabilities. Adding distributed caching to your server farm or compute grid application is now easier - and faster - than ever.

Version 4.0's new features are described in more detail below.

New Support for Linux and Solaris

With production support for the Red Hat Enterprise Linux and Solaris operating systems, version 4.0 dramatically expands your deployment options for ScaleOut StateServer. For maximum flexibility, you can:

• create a distributed cache which mixes Linux, Solaris, and Windows servers,
• mix Linux, Solaris, and Windows clients accessing the same distributed cache,
• access and manage Linux and Solaris hosts from Windows clients, and
• access and manage Windows hosts from Linux or Solaris clients.

The Linux and Solaris versions add two new Java APIs which match their Windows C# counterparts. They also add Web-based management using a PHP-based Web site that gives you the full functionality of the the Windows GUI management console.

Enhanced Communications Architecture

Version 4.0 of ScaleOut StateServer incorporates a redesigned and optimized communications architecture which delivers significantly faster throughput and linear scalability through at least 64 servers. This throughput boost directly translates into measurable performance gains for all applications using ScaleOut StateServer as their distributed caching solution. In recent tests in the Microsoft Enterprise Engineering Center, ScaleOut StateServer was subjected to maximum access load in tests that ramped from 2 to 64 servers, with more than 2.5 gigabytes of cached data and a sustained throughput of over 92,000 accesses per second using a 20 Gbits/second Infiniband network. ScaleOut StateServer provided linear throughput increases at each stage of the test as servers and load were added.

Here is a graphical representation of the test results:

Named Cache APIs

Version 4.0 adds new "Named Cache" APIs which allow developers to easily organize and manage groups of cached objects as collections within a single name space. These APIs take advantage of C#'s indexers to simplify code that accesses an SOSS cache while providing the full capabilities of ScaleOut StateServer's existing "Cached Data Accessor" APIs. (Java support for these APIs on Windows will be available shortly.)

For example, you can now create a name space and add an object to the cache with the following straightforward lines of C# code:

Parallel Query

Version 4.0 also adds the ability to perform fully parallel queries on cached objects. Developers can attach metadata or "tags" to cached objects and query the cache for all matching objects. For example, stock-price objects could be tagged with a ticker symbol or a risk rating, allowing an application to find all relevant stock objects for analysis. For fast, scalable query performance, ScaleOut StateServer performs queries in parallel across all caching servers and employs patent-pending technology to ensure that query operations are both highly available and scalable.

The following figure illustrates how a parallel query is simultaneously executed on all servers and the results combined for delivery to the client:

Mixed 32/64-bit Operation

Using version 4.0, 32-bit and 64-bit caching servers can be mixed within the same server farm. In addition, both 32-bit and 64-bit client applications can access either 32-bit or 64-bit servers, opening up opportunities for greater efficiency and flexibility. For example, a 32-bit Web server application can migrate to a 64-bit server and immediately take advantage of the server's expanded memory for caching application data without the need to immediately port the Web application to 64-bit.

Other Enhancements

In addition to the new features mentioned above, version 4.0 also adds the following new capabilities:

• support for client access timeout so that clients can maintain service level agreements,
• enhanced support for object dependencies so that expiration events are fired when dependent objects are signaled for removal,
• default use of one replica instead of two replicas for each stored object to conserve memory,
• performance tuned to support larger server farms for maximum access performance and faster membership changes,
• new extensibility features for Data Accessor APIs,
• support for unattended installation,
• new join_wait and leave_wait management commands,
• a complete set of performance counters for monitoring host, client, and GeoServer usage, and
• sample code that shows how to make SOSS a second-level cache provider for NHibernate.

Download ScaleOut StateServer evaluation software now!

Review the standard features and benefits of ScaleOut StateServer.