by joe_blaze » July 16th, 2015, 5:04 am
APLNext Application Server is a convenient, high-performance tool to expose APL+Win-based calculations and business rules to web-based clients. APLNext Application Server technology dramatically increases the deployment options for APL+Win based application software. Any client GUI, such as html, pdf, APL+Win, Microsoft Office or .Net WPF, can be easily and efficiently connected to functions in APL+Win workspaces when supported by APLNext Application Server. Dynamic or static content returned to the client to satisfy a request can be in any format including html, pdf, Microsoft Office, etc. APLNext Application Server also supports server-to-server and application-to-server interaction, both synchronously and asynchronously.
Since its public availability (2001), APLNext Application Server has supported automatic multi-threading and request queueing as well as programmer-controlled scaling, load-balancing and client state-selection. Multi-threading implies that when many client requests are received by an APLNext Application Server server, they are each be processed independently. This facility in APLNext Application Server does not depend on the number of server machine processors or 'cores'. Multi-threading is performed automatically without the need for programmer intervention. In addition, APLNext Application Server, using Microsoft .Net-based technology, will automatically take advantage of multiple processors or 'cores' of a web server machine.
As client requests are received by an APL WebServices server they are automatically queued for deployment to available instances of APL WebServices servers.
The number of web server machines in a grid of APL WebServers is unlimited and programmer-controlled. As client request demands increase or decrease, the APL WebServices administrator can scale-up or -down by deploying additional or fewer web server machines as necessary in a cascade structure. APL WebServices automatically performs the allocation of client requests to the cascade of web servers.
Load balancing is supported within the APL WebServices administration tool so that programmer-defined minimum and maximum numbers of APL WebServices instances can be deployed for each application being supported by a web server. Applications with greater priority can be provides with greater resources as needed compared to other applications with lesser priority.
APL WebServices provides for programmer-selected stateless or stateful processing of client requests. In stateful operation the client is bound to a specific APL WebServices instance so that multiple requests from this client can be properly and securely directed to only this client until the binding is released. In stateless operation each client request, from the same or different client, is handled independently.
In all cases instances of an APL WebServices server inherently maintain strict separation and security of different client requests.