Client work stations play a huge role in the performance benchmarking of any JAS architecture! This statement befuddles many people who view JAS as a sole entity entirely dependent upon the web server and the available bandwidth for performance. If type of processor on the client machine is mentioned in the recommendations list, there's a reason for it !!
Document object Model or DOM has a huge role to play in the working of EnterpriseOne JAS. DOM is a set of conventions for interacting with the objects in the HTML, XHTML or XML documents. DOM presents HTML document as a tree structure. Objects in this tree can be accessed and manipulated using methods on them as specified in the DOM APIs.
Now, EnterpriseOne heavily relied on DOM objects for its functioning. The grids, for example, are stored as DOM elements on the client PC. These elements are processed and rendered by Java Script running on this client PC. So the entire time required to fetch data into a grid is spent on the local machine in processing these DOM elements rather than on the Web Server. Clearly larger the number of columns in the grid larger will be the number of DOM elements and hence greater will be the time to fetch, So now, if the grid columns are fixed and the database type is constant, its clear that a dual core pc will take far less time as opposed to a single core one in fetching records in a grid.
This is the main reason of recommending a level of hardware architecture for the local workstations when sizing for a new EOne implementation.
Document object Model or DOM has a huge role to play in the working of EnterpriseOne JAS. DOM is a set of conventions for interacting with the objects in the HTML, XHTML or XML documents. DOM presents HTML document as a tree structure. Objects in this tree can be accessed and manipulated using methods on them as specified in the DOM APIs.
Now, EnterpriseOne heavily relied on DOM objects for its functioning. The grids, for example, are stored as DOM elements on the client PC. These elements are processed and rendered by Java Script running on this client PC. So the entire time required to fetch data into a grid is spent on the local machine in processing these DOM elements rather than on the Web Server. Clearly larger the number of columns in the grid larger will be the number of DOM elements and hence greater will be the time to fetch, So now, if the grid columns are fixed and the database type is constant, its clear that a dual core pc will take far less time as opposed to a single core one in fetching records in a grid.
This is the main reason of recommending a level of hardware architecture for the local workstations when sizing for a new EOne implementation.
