RSS |
 |
| F E A T U R E | |
| |
Mission-Critical Support for NT Failover February 7, 2000 By John Shirely In the growing world of newer, more stringent "high-availability" expectations by network administrators responsible for Microsoft Corp. Windows NT servers and domains, higher standards are being required that formerly applied only to big-iron-class hardware and software. As NT expands further into data centers responsible for 24x7 uptime, more vendors are stepping up to the plate to provide the necessary mission-critical failover solutions. For this article, we tested four of the more prominent players: Computer Associates International's SurviveIT, Microsoft's integrated Cluster Server (included with NT Enterprise 4.0), Legato Software's Co-StandbyServer for Windows NT and NSI Software's Double-Take for Windows NT. SurviveIT won our Editor's Choice award, offering outstanding scalability, connectivity and failover options. It lets you use dissimilar server platforms and components, and can operate over WAN links, enabling a geographically diverse failover plan. Second-place finisher Double-Take goes another route: Rather than failing over applications, it fails over data sets, focusing on data-file replication and mirroring. It, too, works well over a WAN. Unlike Double-Take and SurviveIT, Co-StandbyServer offers very limited scalability--it lacks one-to-many and many-to-one options--but it fails over applications quickly, and reduces the use of network resources by each host machine. Cluster Server also lacks support for many-to-one failover, as well as WAN support, but it alone provides application load-balancing. We chose these products because of the very different approach each takes to the concept of failover and how to properly achieve it. In this article, we cover the main concepts and features you should expect to find in an enterprise-class failover product, including what types of failover modes are used, what type of cluster (if any) is necessary, scalability issues, synchronization methods, WAN support, the various management interfaces and what resources are necessary for a successful installation. In this segment of the disaster-recovery and -avoidance market, the major players are still a long, long way from "five nines" (99.999 percent) of excellence--a category telephone company switching equipment achieved decades ago--but clearly their design goals are heading in that direction. For that kind of ultimate robustness, you have to step up a rung or two on the ladder and move into some very expensive proprietary hardware/ software compilations. Server stress-testing and extreme-to-severe application failover support testing was not a component of our tests for one reason: It would be misleading. Any of these products can successfully replicate your data and fail over a drive, volume, directory or file for you with aplomb; however, actual application failover and performance is a tricky business. All but Double-Take support true application failover, but no one product can do this better than the others for all applications that can be clustered. To test performance parameters by picking only a single application (a SQL database, for example) could put any one of these products at a distinct disadvantage because of the focus of its respective vendor, but that same product might also turn around and give the others a trouncing in a different application category. Additionally, heavy scenario scripting is required to properly fail over an application successfully; if a script isn't optimized properly for a particular application's unique deployment on your network and your cluster, its performance outcome won't show you the whole picture. In other words, your mileage may vary, so don't believe the sticker on the window.
| |
|
PAGE: 1 I 2 I 3 I 4 I 5 I 6 I 7 I 8 I NEXT PAGE |
|
Best of the Web
Data deduplication: Declawing the clones
Data deduplication is emerging as a critically important new arrow in the storage administrator's quiver to answer hard questions about the increasing problem in storage growth costs.
Compression, Encryption, Deduplication, and Replication: Strange Bedfellows
One of the great ironies of storage technology is the inverse relationship between efficiency and security: Adding performance or reducing storage requirements almost always results in reducing the confidentiality, integrity, or availability of a system.
WAN Optimization Whitelists and Blacklists
Optimization is a fantastic way of saving money and creating really happy customers at the same time, but it doesn't work flawlessly for all applications.
WAN Optimization as a Managed Service: It's Not About the Cost
This insight examines how organizations outsourcing their WAN optimization initiatives to a third-party go about achieving their goals for application performance, reducing operational costs, and streamlining enterprise infrastructure.
