The IT view of User Data Management

 Email Article  

Tweet  

  

  

In my last 2 posts (What is User Data? and What is User Data Management?), I tried to define user data and what it encompasses, along with the management requirements that companies want to address.  As I mentioned in my last post: there are TWO stakeholders to user data management: IT and end users.  An effective solution must meet the requirements of both parties or it will lead to dysfunctional behavior in at least one of them.  In this post I'll try and explore the IT view point, with my next post looking at the end user point of view.  

The IT View

From the IT perspective any solution around user data management must take into account the sheer number and distributed nature of end users.  In any company, end users far outnumber IT staff, usually by at least 1-2 orders of magnitude.  Additionally, they are all outside the datacenter and distributed across geographies - further complicating the picture.  It is no wonder that most IT pros shudder at the thought of anything to do with management of data for thousands of distributed users.  In talking to IT pros, they place pretty stringent - but necessary - requirements on end user data management solutions:

• Ease of management: Even a single action that needs to be done once per user or a daily operation for even a fraction of the end user population results in unacceptable and prohibitive administrative overhead. For example, if you're doing laptop backup and if you have to follow up on failed backups for even 5% of your end user population every week - that becomes an unacceptable overhead. If IT has to recover end user data that quickly becomes a very expensive proposition. Or, if you have to periodically perform tests to ensure recoverability of your laptop backup data - that quickly adds up to being prohibitive. Most IT organizations are neither staffed nor inclined to do that.
• Cost per user: The cost per user is extremely important.  The cost is comprised of several items: infrastructure CAPEX and OPEX is obviously important, but there are several other aspects of it like ease of management (see above).  Focusing on just CAPEX and OPEX: numbers which individually seem small can add up pretty quickly.  For example, several consumer focused laptop backup or desktop backup solutions cost $5-$6 per user per month; while that does not seem like a lot for a single PC, when multiplied with 10,000 PCs it amounts to $2.1 M over 3 years - not cheap by any means!
• Control without the burden: IT is expected to be responsible for management of all company data, including user data.  IT pros are expected to have control of the user data and be able to answer yes to the following questions:
• Can we recover data if it gets deleted?
• Do we know if John in payroll has downloaded the entire payroll database on his laptop?
• Can we rapidly figure out what data was stored on the stolen laptop to assess our risk exposure?

IT is expected to and wants to do the above, but without having to incur prohibitive burden on already thin staffing.

• Integration with existing user management systems: Integration into existing end user management systems like Active Directory for single sign-on and software deployment are critical - otherwise there are more systems to manage and keep track of the multitude of users in an enterprise.  Taking the example of cloud based laptop backup and desktop backup solutions, most of them have a problem here.  They typically don't integrate into existing Active Directory deployments requiring IT to manage yet another set of user credentials, which is not only a pain, but also a security concern.

Faced with above requirements, storing data on laptops seems like an anathema - because there are few solutions out there that allow IT the level of control it needs for laptop data at the cost point and ease of management that's needed to meet the above requirements.  So, naturally, IT gravitates towards requiring that all user data be stored centrally within the datacenter(s) on Network Based Shares or to use the technical term: Home Directories.  However, today's end users are quite demanding.  They are increasingly working disconnected and are accessing their data using an ever increasing medley of devices (Blackberry, iPhone, iPad anyone?).  The central data storage technologies of network based shares were invented in the early 90s and simply don't work for the access profiles of today's users.  As a result, the end users are silently revolting against the IT requirement to store all data centrally.  End users are not only storing confidential company data on their laptops, they are doing all kinds of crazy stuff to ensure that their data meets the accessibility patterns they need - often putting companies afoul of regulatory requirements, so there is a clear disconnect here.  I'll explore this in my next blog.