Leave feedback

Persuasion Infrastructure

Architecture and platforms


StreamServer’s architecture is service based, multi-threaded, and distributed. Based on a platform-independent architecture, the solution can be deployed on either Windows or UNIX platforms. This multi-threading approach allows optimal usage of multi-processor servers. With its high-performance streamlined architecture, it only requires minimal hardware resources.
The architecture of StreamServer and its adjacent components are optimized for integration within an enterprise IT infrastructure. StreamServe Persuasion provides the means to fulfill critical enterprise requirements including scalability, reliability, performance, security, and extensibility.

StreamServer

  • Microsoft Windows XP Professional – Service pack 3
  • Microsoft Windows 2003 Server Standard Edition – Service pack 2 and later
  • Microsoft Windows 2003 Server Enterprise Edition – Service pack 2 and later
  • Sun Solaris 10
  • IBM AIX 5.3.x
  • IBM AIX 6.1.x
  • HP-UX 11i V2
  • HP-UX 11i V3
  • SUSE Linux Enterprise Server 10
  • Red Hat Enterprise Linux ES/AS 5

Design Center

  • Microsoft Windows XP Professional – Service pack 3
  • Microsoft Windows 2003 Standard Edition – Service pack 2 and later
  • Microsoft Windows 2003 Server Enterprise Edition – Service pack 2 and later
  • Microsoft Windows Vista Business

Control Center

  • Microsoft Windows XP Professional – Service pack 3
  • Microsoft Windows 2003 Standard Edition – Service pack 2 and later
  • Microsoft Windows 2003 Server Enterprise Edition – Service pack 2 and later
  • Microsoft Windows Vista Business

Reliability and Scalability

StreamServe Persuasion contains a number of mechanisms that enable the distribution of crucial system functionality to multiple server instances. This provides scalability, load sharing, geographical distribution and resilience. For example, input data can be buffered within a central and instance- independent input queue before the StreamServer processes it. Individual server instances can be taken out of the framework for maintenance or other reasons without production downtime.
Other servers in the network can seamlessly manage the load of the affected instance, so that availability is unaffected. The use of an independent input queue can be deployed within an availability/failover scenario to correct downtime associated with hardware malfunctions, network malfunctions and/or repository malfunctions. Dedicated servers can be configured to handle specific document types requiring immediate processing such as invoices and POs. In the event that servers become unavailable, workload can be dynamically redistributed to alternate active systems ensuring increased resilience in failover situations.
It is also possible to apply load balancing to a StreamServer deployment. If too many jobs for a single instance arise the system will automatically provide load sharing between the available instances. Multiple processors distributed across multiple servers can be grouped together as logical resources.
Advanced queue management ensures workload gets dispatched in accordance with business demands. Decoupling the processing of incoming business documents from the production of output data streams enables increased use of multiple processing threads. High volume output, produced by billing systems and other enterprise applications, can start to be printed sooner with StreamServe Persuasion, ensuring increased throughput and improved utilization of hardware resources.
Due to this architecture, StreamServe Persuasion meets high demands on scalability, availability and reliability. A strong, solid and scalable system architecture is key to the achievement of an enterprise-class product like StreamServe Persuasion. StreamServe Persuasion is well suited to large, critical environments and can scale with the growth of an enterprise.
Within server instances, multi-threading capability and queuing enable parallel processing of as many threads as there are processors available. In multi-server architectures, capacity can be added flexibly when and where required. If the existing installation reaches capacity limits because of increasing document volumes, installing additional servers can easily extend capacity. With the solution’s modular architecture, this is possible at any time and with minimal effort.

End-to-End Tracking

Feedback on the successful delivery of documents can be provided to the originating business application even when a third party application, such as an output management system (OMS) is used for output delivery. Status information received from the external OMS is associated with the application request and communicated back to the originating system. Customers benefit from end-to-end tracking of their presentment even when they are processed by multiple systems, thus making it easier to pinpoint and resolve problems.

Performance

The performance of the StreamServer depends on the processes (calculations, sorting operations, etc.) and the contents (graphics, logos, etc.) as well as the configuration of the platform where it is installed.
A typical StreamServe Persuasion installation processes thousands of pages per minute per CPU. Under optimal conditions, the StreamServer can acheive up to 100.000 pages/minute/CPU. Since StreamServer uses multi-threading, processing power can be arbitrarily increased by adding additional CPU power to the underlying hardware. The architecture of StreamServer and StreamServe Persuasion has been oriented towards delivering superior performance for any requirement. The result is a universal solution, of which performance is not a fixed number, but a scalable variable that can be adjusted as needed.
The StreamServer fulfils all preconditions required for production printing. StreamServe Persuasion enables the splitting of documents, sorting and conversion. In truly large scale printing environments, specific tasks, such as sorting, can be assigned to dedicated hardware resources, which greatly benefit overall processing performance and make the system perfectly scalable.