Introduction to Policy-Based File Area Network

Sixteen years ago, TNS had only a few PCs (Personal Computers)—a handful of people had them, most did not. Now, just about everyone in our offices has a PC, and it’s not uncommon for some of us to use more than one at the same time.  Initially, files were shared using floppy disks, but this eventually became very impractical.  New ways to network PCs and share files were discovered, and we also introduced what we call a File Server to store everyone’s files, so sharing files became much easier.  Pretty soon however, we needed more File Servers, one or more for each function or department. 

As we go further into this digital world, digitizing anything and everything possible, we create hundreds and thousands of files scattered in hundreds of millions of PCs and servers, resulting in more and more storage across many different types of network architectures.  For most computer users and organizations, our digital information is becoming a lot like our garages before Spring cleaning: for some, the garage just keeps getting crammed with more and more stuff, and there is never any Spring cleaning because it’s just easier to leave it alone than to face the daunting task of trying to sort through all the junk.

Because of our own creation, finding files and information within these undisciplined and unstructured mountains of data is increasingly difficult.  With or without our realizing it, we are at a tipping point now: we need a better answer to our increasing volume of digital files.  Many people are trying to provide solutions to this issue by using search engines—faster and better search engines built into Operating Systems, but the search approaches are just a short-term patch and do not address the root causes—the conditions that have created the problem in the first place.

So, what would a new and more effective method of file storage look like, and how would it be better than what we already have?  What we have today is a totally unstructured way of storing files anywhere we can find space for them, and most of the time, we do not even fully understand where we are saving the files, making it difficult to find them later. To compound the problem, most of the files we are saving eventually become obsolete or otherwise irrelevant, but nonetheless continue to take up disk space. We save various versions and duplicates of files that are outdated or irrelevant and never delete them.  That occurs over and over again because we have no way to identify the files that we need from the ones we do not. 

Policy-Based File Area Network

Above is our proposed high-level diagram of a possible solution.  To my knowledge, there is little to no evidence that we are anywhere close to solving this issue.  However, there are a few solutions in the works that are attempting to address the issue. The most widely available but not very well known solution is Microsoft’s Distributed File System (DFS), built into the Windows 2000 & 2003 Operating System.

Microsoft’s DFS lays the foundation for what could be richer features, functionalities, and better architecture to come.  It focuses on virtualization of the file storage, provides fail-over and load-balance in a Wide Area Network (WAN) environment, as well as some ability to centrally provision and manage file storage architecture.  Unfortunately, the existing DFS still lacks many essential features and functionalities that will make files more accessible and manageable, quickly and transparently for users and IT alike.

We need features and functionalities in the area of governmental and organizational regulatory compliance filters, ways to eliminate duplications, version control, file-type cost control, network-wide instant keyword and content search engines, and more sophisticated file migration and archiving, and we all need to comply with governmental and organizational policies in this area.

By Benson Yeung, Senior Partner

Back to Top