In the last several years, managing IT systems has become a major challenge for administrators. Most IT departments are stretched to the limit, as they contend with exploding volumes of data, proliferating systems and applications, and an ever more complex operating environment.

IT shops tend to fall into two camps: smaller or less sophisticated groups patch and fix as they go, while more proactive organizations actively use systems management tools. Larger IT groups may have a tool to manage every piece of hardware—from network equipment and servers to switches and routers—as well as software management solutions.

Mired in daily maintenance

As enterprises scale out, managing an exploding number of hardware devices can become a nightmare. Updating and patching drivers, operating systems and applications can be a real challenge, not to mention monitoring systems for performance and capacity. Most IT shops find themselves mired in daily management tasks, finding it difficult to focus on the strategic issues that add value to the business.

To some extent, the move to open standards can complicate systems management. Because most vendors include or sell management tools with their hardware, it’s not unusual for data centers to run dozens of these tools concurrently. In addition, many data centers have comprehensive management consoles from industry leaders like Microsoft, Novell and Altiris, which add another layer of complexity.

“IT staff frequently spends a lot of time managing their management applications,” says Ed Reynolds, Senior Manager for Dell Systems Management Marketing. “Their big challenge is to keep current with all their software revisions and patches while getting a single view of everything that’s going on in their IT environment.”

Standards keep things simple

But because standards seek to minimize the number of interfaces and tools required to monitor systems and keep performance at required service levels, they offer a roadmap to simpler and more effective systems management. Vendors like Dell are driving the standardization of the instrumentation needed to monitor hardware so it can be collocated with the same agents that monitor operating systems and applications.

With integrated hardware and software systems management, enterprises will have more control over their IT environments and be able to troubleshoot problems because all data resides in the same location. For enterprises that want to implement proactive monitoring systems to handle the growing volume of incoming information and alerts, it makes sense for IT managers to look at the four main stages of enterprise systems management.

Where are you now?

• Stage 1: User-driven monitoring . This is the reactive state in which many data centers operate. End users call in when there’s a problem, and the IT department responds. To move beyond this, instrument your systems and learn the underlying factors that affect uptime.
• Stage 2: Service-level monitoring. By using instrumentation to monitor service levels, you can receive proactive failure alerts that tie into specific hardware, such as a disk or network controller. You also gain visibility into network and CPU utilization, enabling you to identify potential causes for failures, such as high utilization or a rise in temperature.
• Stage 3: Proactive systems management. At this stage, you can track system performance and take proactive steps to avoid system failures and saturation points. As you collect information, you can use it to make smart decisions in allocating resources, capacity planning and anticipating over-utilization. Because you’re no longer in reactive mode, you can now anticipate whether systems need servicing or new equipment needs provisioning before problems develop. And if problems develop, a notification can be sent to the vendor, a replacement part ordered, and a service call scheduled, all automatically.
• Stage 4: Policy-based management. At this stage, you’re using systems management strategically by creating policies that take defined steps to respond to certain events and conditions. In virtualized environments, a typical example of policy-based management might be that, when a usage threshold in an application server farm is exceeded, policies are in place to automatically provision a new virtual machine and redistribute traffic among virtual machines, transparently handling the load. At this stage, hardware and software management is integrated, allowing you to use policies from your choice of enterprise management tools.

While policy-based management might seem a distant goal for IT shops that are just reacting to problems, virtually every enterprise can benefit from trying to achieve a more proactive systems management environment. Look for opportunities to reduce the number of tools in use and take advantage of the new instrumentation that industry groups like the Distributed Management Task Force (DMTF) are developing.

In our next article on industry-standards systems management, we’ll explore two ways to move ahead with management tools and the promise of policy based management.