IT professionals must look beyond cost and consider five major areas when selecting a systems management tool for the data center.
IT professionals can no longer rely on traditional physical servers and provisioning to manage enterprise computing. Today, server virtualization and other technologies abstract workloads from underlying systems; mission-critical applications may reside on almost any server at any location, connected through a network that is also virtualized to manage traffic flows and data. IT professionals need powerful tools to manage such systems. The array of available systems management tools and their feature list is growing, so how do you choose the right one?
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
1. The required feature set
The search for a system management tool must begin with a wish list of features and functionality. This quickly narrows the field of products by eliminating tools with inadequate or superfluous features -- which can save considerable expense. You can always upgrade later.
Some possible features include:
- Hardware discovery, inventory and configuration information, allowing administrators to see hardware-level details -- temperature, fan status and other metrics -- of the various servers, storage, network and endpoint devices located across the data center.
- Software management for hypervisors, operating systems and applications that can inventory the software in use and identify versions. Some tools use that information to automate software updates and patches across the data center, keeping unproven software from disrupting production systems and lifting a workload from administrators' shoulders.
- Notification systems that warn administrators about troublesome situations. Reporting features can summarize various issues and provide detailed, actionable information.
- Lights-out management, which uses a dedicated management network to manage and monitor systems even when they are powered off.
- Security features that can establish and enforce data center security using configurable policies.
- Service desk features that provide ticketing and workflow capabilities so IT staff can centralize, prioritize and track the state of any reported problem or user support request.
- Automation capabilities to help busy administrators handle routine management tasks across large numbers of systems while minimizing errors.
2. Heterogeneous or homogeneous systems support?
Remember that for a tool to manage a system, it must be able to "see" inside the system to gather data and exercise control. This is simple for homogeneous environments that use similar server architectures. In fact, major server vendors offer comprehensive systems management software, such as HP OpenView, IBM Tivoli Framework and Dell OpenManage, already tailored to the vendors' hardware families.
Heterogeneous environments present a serious challenge -- a single tool may not be able to gather the same data or impose the same levels of control on every make and model of system hardware. Heterogeneous tools often sacrifice some granularity to broaden the scope of supported systems. Data-gathering agent software modules installed on each managed system can bolster capabilities, but agents also risk slowing system or network performance and must be updated and patched periodically.
A tool with all your must-have features has to also work across the entire spectrum of hardware currently or potentially deployed in your data center. For example, if you're depending on a tool to provide system inventory data, the tool will be useless if it only inventories certain systems and not others.
3. Open source versus proprietary tools
In addition to the system management tools from major hardware vendors, there are many third-party vendors in the market: Kaseya, Zyrion, ManageEngine, Hyperic and even Microsoft. It might be worth your time to evaluate open source system management tools: Nagios, OpenNMS, Puppet, Spacewalk, Zenoss Core and others. Well-respected open source tools are usually on par functionally with proprietary counterparts -- at a lower price.
Open source development is commonplace in Linux environments where code is freely available, guided by established open source licensing agreements. Developments and changes to the product are freely available and driven by the user and developer community. However, such development can be slow and technical support spotty, if it exists at all. Some larger open source developers provide both open source and enterprise versions of products, where the added cost gets you enterprise-class support.
4. Integration with other tools or suites
Large enterprises have their pick of comprehensive software suites that do just about every imaginable systems management task. Smaller enterprises often adopt a set of individual tools that each operate somewhere within the spectrum of systems management.
Unfortunately, the data generated by one tool may not be fully compatible with others. The resulting integration problems mean that multiple tools cannot effectively share their unique insights through a single interface. IT professionals must learn and use tools individually, so the risk of management errors or omissions because of missed information shoots up.
When evaluating any tool, find out how it interacts with other systems management tools in your environment. There is no substitute for testing. For example, if one tool is already storing detailed system hardware and software inventories and configuration data into a SQL database, a new tool used to automate system setups and provisioning should access that database. If not, the new tool is far less valuable -- someone would need to provide it that data from scratch.
5. Installation footprint and learning curve
Every system management tool requires computing resources: a server, storage and the availability of a database. Simple solutions may have very light requirements, while comprehensive management suites impose onerous computing demands that scale up with the number of systems being managed. Make sure the business can handle your tools' current and future resource drain. IT will have to accommodate new management modules, more management tasks and a larger base of managed systems in the foreseeable future.
The IT staff also requires some training on systems management tools. There is certainly no substitute for hands-on experience, but sophisticated tools may demand some online or even in-person training -- particularly for large, complex, highly customizable management suites. Consider the costs of training staff, including the ongoing training budget as IT personnel change roles and newcomers join, when choosing a tool or array of tools.
Choosing systems management tools is a challenging exercise for any business. The best path to success considers management needs and business limitations, then matches those to the best proprietary or open source feature set that interfaces to existing management products.