Server Management Tools
What’s interesting about server management, and the tools that support such activities, is the degree of interrelatedness between this topic and network management. In fact, many of the topics defined within the ISO network management model are every bit as relevant to managing servers as they are to managing networks.
Because the ISO model was developed in the 1980s, ideas about what it means to manage networks–and servers–have changed, but its original elements remain relevant. This means additions to and enhancements of the model’s original five subject areas, which include fault management, configuration management, accounting management, performance management and security management. The most obvious additions have to include storage management, directory services management, software distribution for server deployment and patch/update management for server maintenance and security management purposes. Some might argue that these additions are simply elaborations within the area of configuration management, but because all have their own sets of associated tools, they’re worth mentioning as separate areas of specialization.
Classes of Server Management Tools
Various categories or classes of tools cover the server management space. At the top of a hierarchy of systems that includes server management capabilities, you’ll find general-purpose, enterprise-class tools designed to manage networking gear, servers, clients, network infrastructure and much, much more. This includes systems like Tivoli’s Enterprise Console, HP OpenView and CA-UniCenter. All of these kinds of platforms require a considerable degree of customization and placement of agents or sensors to do their jobs. In addition, all of them support standard APIs and include their own proprietary APIs so that a wide range of third-party tools of all kinds–including server management tools–may be integrated under their purview, for alerting and monitoring, and sometimes even their control.
In the middle of this hierarchy, you’ll find various systems designed to provide centralized management and control specifically over the servers in an enterprise or organization. Sometimes, such tools are platform-specific–for example, Microsoft’s Systems Management Server (SMS) or Sun Management Center–while others are designed to accommodate Windows, Linux or UNIX, and other platforms as well. For example, Novell’s recent push into Linux, along with its historical involvement with NetWare, equip it well to deal with multiple platforms. That’s why Novell’s ZenWork and eDirectory products can claim to support Windows, Linux, NetWare and Solaris, all with equal facility.
At the bottom of the hierarchy, you’ll find tools designed for use on individual servers, often as part of whatever underlying operating system is in use. Windows, Linux, Solaris, NetWare, MacOS X–you name it–such platforms invariably include lots of tools and consoles for server management, most of which may be operated remotely or locally. At some level, in fact, most of the items higher in the server management hierarchy inevitably end up working with these components as well, simply because they delegate the final activities and execution of instructions to local facilities to get things done on each individual server that’s managed. Such tools are occasionally necessary when interaction with a single server is required. This is where the finest granularity of control and the highest level of detail are available.
A Case for Specialized Services
Directory services have a special role to play when it comes to server management, access controls, configuration, services delivered and more. That’s because various aspects of an organization’s policy decisions–for example, as they relate to security matters, user rights and privileges, applications, services and protocols permitted or restricted–are often most clearly articulated and easiest to implement and control in the form of directory service settings and data. Some directory services–like Novell Directory Services (NDS), Microsoft Active Directory (AD) or Sun Directory Services–take on important roles related to how servers are incorporated, used and handled within an organization.
Task-oriented, distributed management or monitoring tools built into operating systems, like the Microsoft Performance Monitor and similar management console snap-ins, also fit into this category. The modus operandi at work here might be best described as, “pick a platform or an application, then look for a suitable monitoring tool.”
As you’d expect, lots of these tools are available, so those most likely to be of interest depend on what kind of server or servers you seek to monitor. Use your favorite search engine, look for trade or professional reviews and check out relevant news or user groups to get pointers to equivalents in your areas of concern.
Other bits and pieces of the server management puzzle that are likely to fall inside their own toolsets include the following:
- Server Construction and Deployment Tools: Some organizations wouldn’t consider deploying servers without using Microsoft’s sysprep and remote installation services (RIS); others prefer to create images using tools like Symantec Ghost, coupled with a software distribution server of some kind. Whatever the individual choices in this category might be, the driving notion here is that some method to specify what goes into a particular server’s software makeup and to drive its detailed configuration is needed, as well as a method to ensure its proper installation on one or more target machines.
- Software Update and Patch Management: In an era where “zero-day” security threats are emerging (that is, exploits that depend on newly discovered vulnerabilities occur at the same time those discoveries are made), mechanisms for capturing, testing and distributing patches and updates are becoming increasingly important for servers and clients alike. Lots of solutions exist in this space, most of which tend to be platform-dependent and many of which use secure Web services to enable servers to push, or clients to pull, updates and patches as needed.
- Storage Management: When a server’s responsibilities include providing access to large amounts of storage, perhaps for data-intensive database, GIS or modeling applications, it sometimes makes sense for organizations to centralize storage and deliver it as a service to the very servers that manage its access for the client requesting their services and the applications they use to deliver them. This lets implementers realize massive economies of scale and often provides better reliability, fault tolerance and failover capabilities than would be affordable on a per-server basis. Though higher costs and complexity offset these advantages to some extent, there’s no denying that working with network attached storage (NAS) or storage area networking (SAN) also means working with specialized management tools for these environments.
- Security Management: Certain classes of servers–Web servers, for example–carry specific security profiles and present distinct security management requirements and needs. All kinds of security management tools include server management, monitoring and even scanning and auditing capabilities. For servers exposed to the Internet, it’s essential that such tools be properly, effectively and regularly used.
- Service Metrics: Given that a server’s primary job is to deliver services to clients, it makes very good sense to measure how good a job any server is doing. To some extent, this can be addressed by relatively standard performance metrics, throughput observations, transaction processing rate analysis and so forth. But more and more organizations are turning to load simulators or outside monitoring companies to keep constant tabs on how their serv