Originally posted by Mike Boota:
But what if there are many users accessing the file and do I have to load it in memory for each request or have it loaded in memory once on server startup.
As already suggested , you can have the XML read onto a DOM and then use XPath to get the required values.Once you load it , all the XPath queries can be directed to the same DOM.No need to load it again and again.
One thing to note is , for a file of about 12 MB it would atleast take 18 MB (might be a few more MB's) of RAM.
So if the xml has some information specific to user and in production you will have many more users.In that case the size of the xml would be big and will take a lot of memory.Then you might have to think of something else.