RSS
  By Scott Mason Safeguarding Network Data: This Is A RAID When I have to make the complex decision of choosing a RAID system, I try to narrow the field by finding out which systems support my operating system, run on my servers and have the capacity I need. You can do the same by consulting the Buyer's Guide charts included here. As you look through the charts, you'll be able to quickly eliminate products that don't meet your basic needs. And once you've narrowed the field, you can pose the tough questions to the vendors whose systems are left on your list. In asking these question s, don't think in terms of the technology; think in terms of the business. Which applications will you be running on your server? How critical is the data to your business? How critical is availability? Can you keep the system running through the day and repair it at night, or do you need 24x7 availability? What effect does downtime have on your business? You should base your decision on four criteria: performance, cost, reliability and capacity. Once you're satisfied that you've addressed all your business concerns, finding a RAID system that meets your requirements should be relatively easy. Try to Be LevelHeaded Figuring out which levels are supported and which level you'll need is a good place to start. Put some serious thought into this because the RAID level affects performance, cost, reliability and capacity. Try to get "value" from your value-added reseller (VAR) by getting his/her expert advice on this. And try to avoid getting caught up in all the hoopla about RAID levels. I t's easy for storage vendors to market RAID 10, RAID 30 or RAID 50--which are simply mirrored versions of the usual RAID implementations--in an attempt to differentiate their products. How important is the RAID level? Joe Molina, chairman of the RAID Advisory Board (RAB), says that as far as RAB is concerned, the decision to duplex or stripe your data is only one aspect of data integrity. You have to look at all the components of the total system and ensure that each meets the level of reliability you need. Having redundant controllers in a server with only one power supply is like having two spares in the trunk of a car about to run out of gas. Try to balance all the components in the system. RAB's Web site (www.raidadvisory.com) offers a new storage classification system, with 21 criteria for determining if a system is failure-resistant, failure-tolerant or disaster-tolerant. The criteria range from the ability to regenerate the data from a backup to having mirrored systems with redundant subsystems in remote locations. RAB is trying to get vendors and users to focus on the total system and not just the RAID level. Don't think that the RAID level decision is unimportant. Duplexing will be more reliable, but it is costly when you use more than four drives. Striping data in a RAID 5 array is not as reliable as a duplexed sy stem, but you get better performance when the data is being read from or written to multiple spindles. In most cases, your application will determine the RAID level to configure. Tolerant to a Fault Next consider which features are included in the system. Here you'll need to balance reliability against the cost of keeping the system up and running. If yours is an online transaction processing (OLTP) system, for example, high reliability is critical. You'll want an array enclosure like Storage Dimensions' SuperFlex 5000 with redundant hot-pluggable power supplies, fans and controllers. On the other hand, in applications like rendering farms or prepress, the proc ess is more important than the data. This type of system would be designed for performance rather than reliability. Data loss is not as critical as getting the data moved onto the disk reliably and quickly. Here you would use RAID 1 with lots of drives. The more spindles you write simultaneously, the faster you'll be able to move your data. And you won't waste time reading and writing parity information. In any case, be sure to include system management software that will warn you if a component is about to fail.
|
|
Updated June 6, 1997 |
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.
