NEI (News - Alert) offers a series of platforms that are used both in-house and in the field. The reliability of this technology is based on the technical support calls that are received on any systems deployed and currently in operation. This recent NEI white paper explores the reliability of appliance deployment within the business environment.
When examining configuration, the most basic aspects to determine reliability are the physical hardware and configuration of the server. The specific hardware and operating system configurations in place must provide the necessary server platform and this platform must be tested, verified, and validated to perform at the anticipated level for a specified amount of time.
In any appliance deployment from NEI, common elements include redundant, hot swap power with failure alarming, and sufficient power to provide enough amperage for normal operation of all system components.
Customer-defined configurations are also tested and certified by NEI. Such systems leverage appliance deployment and tend to experience a large amount of operational time in a particular environment. As a result, dependable demonstrated reliability data is provided. These platforms can also offer a specific level of reliability, assuming the essential configuration is not altered.
Reliability ratings in appliance deployment will be affected by how a server is used and how its downtime is measured. Server downtime should be measured according to the definitions and rules outlined in the Revision 3.6 of the TL9000 Measurements Handbook for supplier attributable outages.
In any type of appliance deployment, a system outage would correspond to a server with a “Critical Problem” as defined in the NEI Reliability Report. Technical support and IT departments record critical problems that tend to be directly related to the particular applications installed on systems, the major function of the server and the administration of those applications.
To ensure reliability of any appliance deployment, it is important to not install untested, uncertified or unsupported applications on a server that has been pre-configured; IT should also not install third party utilities, programs or device drivers that directly affect hardware; avoid installing multiple applications on a server for the purpose of utilizing that server for multiple major functions; and perform sound administrative techniques to minimize server downtime.
The overall reliability of the unit is dependent upon good server maintenance and if a redundant power supply or RAID drive fails and is not adequately replaced, the possibility of system failures increases. It is also important to pay attention to the physical environment in which the server will reside. In short, demonstrated system reliability is an estimate of the potential and likely performance of a system based on past performance.
Susan J. Campbell is a contributing editor for TMCnet and has also written for eastbiz.com. To read more of Susan’s articles, please visit her columnist page.
Edited by Jamie Epstein