This tip is the second in a three-part series on remote server management. Read the third tip on best practices for remote server management. You can also read part one about exploring remote server management options.
Remote server administration tools give IT professionals the ability manage data centers without actually being there. However, remote server monitoring–as with any technology–has limitations that should not be overlooked. This tip explains what criteria to consider when selecting remote server administration tools, and offers suggestions for how to best utilize those tools in a data center.
Options for remote server tools
Latency remains a significant issue, and any connectivity problems between the servers and administrator can make remote management slow or even impossible. Yet experts generally agree that the various technologies used for remote server management are relatively mature and robust today. Tools can often provide warnings and alerts that allow administrators to take action before a server failure actually occurs. But confidence in the overall state of remote server administration toolsis where the agreement ends. The actual choice of tools can be contentious.
For example, some experts prefer software tools like Remote Desktop Protocol in Windows for basic server maintenance tasks. Other experts say that integrated solutions like HP Integrated Lights-Out (iLO) and Dell Remote Access Controller (DRAC) cards may be better because they’re bundled with their respective servers and covered under the servers’ warranty. Still, others opt for third-party remote server administration tools because of their ability to gather and present a broader perspective of the remote environment. Similarly, KVM over IP can be a powerful management platform regardless of the operating system or data center environment, but it may not be appropriate for every situation.
The choice of management tools is really minor. The bigger problems with remote server management are actually attributed to people and their response times rather than technology. “Anytime you’re dealing with something that you can’t walk up to and touch, you’re at the mercy of other people,” said Ian Parker, senior Web services administrator for Thomson-Reuters, a business intelligence company based in Ann Arbor, Mich. Other people’s schedules, differing work priorities and lapses in interpersonal communication can make delays in response time seem more frustrating or painful than they might actually be, added Parker.
In many cases, sound policies and measurable procedures can help to minimize skewed perceptions and ensure that on-site personnel address server issues in a timely manner. If remote service is provided by a third party, chances are that the service contract or service-level agreement already incorporates a well defined escalation path and measurable response time windows.
Perhaps the biggest gap in remote server management technologies is in automation. For example, Parker said that many of his remote servers experience extended idle time during off-peak hours. Better automation support from remote server administration tools might allow organizations to migrate more idle VMs onto fewer physical hosts, power down the additional physical servers during idle hours and then power up those servers and reallocate the VMs in preparation of peak demand hours. Parker said the idea is there, but more development is needed in that aspect of remote server management products.
This was first published in August 2011