The Peer to Peer model lets users access database content stored on several server locations without any of the backend architecture being exposed. This example demonstrates a system which prepares a customised report (eg. a pricelist) from a remote server and makes it available as a download immediately after it is generated.
Description
WebPublisher Studio can be used to arbitrate requests to and from any TCP/IP capable server. This also means that one WebPublisher server instance can act as a client interacting with another instance of WebPublisher server to return data. It also provides a service-specific firewall as well as allowing a data bridge to databases that are located on different servers or discontiguous networks.
How it works
In this configuration there are two instances of WebPublisher Studio located on separate machines on different networks. Requests from a web browser are routed through the first instance of WebPublisher Studio server which then sends a request to a second instance that has access privileges for (1) an FTP server and (2) access to a SQL database storing enterprise data.
This second instance runs a report on the SQL database over ODBC and outputs it as a PDF file into a repository on the FTP server. It then replies to the first (requesting) instance with a hyperlink URI to the document it has created. The first instance then embeds this hyperlink into an active server page and returns this page to the web browser.
The user can now access the report over HTTP by clicking the embedded hyperlink to the PDF file on the FTP server.
We use a peer-to-peer system in the proPIP web-service application we wrote.