Separate the Administration Tool from a Web Server
IntroductionIn some organisations there are strict network protocol restricting connections between computers. This knowledge base article describes one such situation and how it is overcome using the Triaster Platform. The situation which this article describes is illustrated below: ![]() There are two issues with the above setup that make it incompatible with our standard configuration, these are:
The solution we found was that we hosted the Browser Toolkit Administrative Tool on the Application Server and this article will outline this procedure. ProcedureThe tasks need to be completed are:
These are described in more detail below. Install a Publication Server and Process Navigator on to the Application ServerThis is as per a normal install except that you will need to publish the ProcessLibraries content locally and the URL will be the URL of the Web Server. Install a Browser Toolkit and a Process Library on to the Web ServerThis is almost as per a normal installation. There is no need to configure this Process Library as it will be configured on the Publication Server. Please note that the Process Library needs to be installed in to the same folder as the Browser Toolkit. Install a Browser Toolkit and a Process Library on to the Application ServerThis is slightly different to a normal installation. The Browser Toolkit installation is as per usual, also please note that the Process Library needs to be installed in to the same folder as the Browser Toolkit and these further instructions should be implemented within the installation structure: There are 3 types of files that need to change in the Application Servers "ProcessLibraries" folder, these are the "Libraries.xml" and the "Frames Config.xml" contained within the "ProcessLibraries" root folder, and the frames.html in each Sites "Tools/Admin" folder. Details of these are as shown below.
Write a Script to transfer the files from the Application Server to the Web ServerNow while the Application and Web Server cannot see each other there does need to be a computer that can see them both that can transfer the published content. This content can therefore be transferred using an xcopy, a robocopy or possibly even using FTP protocol. This would need to be run as a scheduled task every time you would want the library content updated. An example of an xcopy script is shown below: xcopy "\\PublicationServer\Triaster\ProcessLibraries" "\\WebServer\triaster\ProcessLibraries" /i /s /d /y |