A Google search on software audits will yield many articles that point to a dramatic rise in audit activity in recent years as software vendors look to cover the shortfall in conventional revenue. Many of the articles are published by or sponsored by organizations that offer tools that help with software asset management; a discipline employed to help companies prepare for software audits. What is puzzling is that these tools, and there are a great many of them, are almost exclusively focused on the Windows desktop environment when the majority of software audits are undertaken by companies that sell server software. It’s all well and good knowing exactly how many copies of Adobe CS5 or Microsoft Project you have installed, but when Oracle, IBM, HP etc comes knocking that data is of no use whatsoever.

There are many reasons why this paradox exists, certainly more than can be covered in a blog post, so I’ll attempt to focus on the reasons that are most relevant. In relative terms, desktop SAM tools are easy to develop. Windows, Mac and Linux workstations have well-defined interfaces that can be used to identify the applications and executables installed upon them. Numerous external catalogs exists that enable the identification of the licensable packages that discovered applications and executables represent, and license metrics fall into only a handful of well-understood classes, such as seat, CAL, named user etc. Desktop SAM tools differentiate themselves by layering on top of this basic information, facilities to make the management of entitlements easy, and adaptable to organizational processes; such as departmental allocation.

In the server realm, no such standardization exists; every vendor implements licensing differently. Whereas in the desktop world the same methodology will work just as well for Adobe as it does from Microsoft (notwithstanding small reporting and packaging idiosyncrasies), in the world of the server, Oracle licensing will have virtually nothing in common with IBM. Each of the major vendors requires a different approach to measuring license consumption for each of its major product lines. For example, license consumption for IBM’s DB2 on a mainframe, will use a totally different model to the measurement of license consumption for WebSphere ESB on a P-Series server. (For more information on the different types of IBM licensing model, an excellent article can be found here). The SAM tooling required to automate this calculation tends to be vendor and product-line specific and subject to customization.  Imagine if your desktop SAM tool only supported Microsoft Office, and you needed a different tool (or an add-on) to support Microsoft Project. This level of complexity has resulted in fewer vendors offering tools for server SAM, which in turn leads to the tools being more expensive.

SAM has long suffered from a lack of upper-management support; the tools are viewed as existing to merely make the job of software procurement easier. As desktop SAM tools have come under increasing price pressure, the risk-reward equation has made the market more buoyant, however server SAM benefits from no such broader market shifts. Consequently server SAM remains the preserve of those organizations that take a strategic view of SAM and are willing to invest to support a more professional approach to managing licenses and eliminate the pain and organizational disruption that server audits always cause.