| TABLE OF CONTENTS
Boosting RAID performance
- RAID-53: RAID by any other name
The first thing to understand about RAID-53 is that it has the wrong name. A better name for it would probably be "RAID-03," as it involves, in essence, striping (RAID-0) a series of RAID-3 (striping with a separate parity disk) arrays without additional parity.
- RAID form factors shrink as market grows
The RAID market has continued to grow at an impressive rate, despite the relative lack of new developments in the technology itself.
- RAID-50: RAID-5 with suspenders
Like the other double-digit RAID levels, RAID-50 is a combination of two basic RAID techniques. It combines striping (RAID-0) with independent data disks with distributed parity (RAID-5).
- RAID-3: What is it good for?
You don't hear much about RAID-3, but it has its uses. Find out when to use RAID-3 and why.
- RAID-1 vs. RAID-5
RAID-5 performance is notoriously slow for writes, especially small writes, because of the CPU overhead that each write requires.
- Glossary term -- RAID
Find out exactly what RAID means.
- The essential RAID primer
Always wanted to be a RAID know-it-all? Take a look at part one of Evan Marcus' RAID primer where he covers the various levels of RAID. Each level's description is concise and to the point. Bookmark Evan's primer and brush up on the RAID.
- RAID-10: Two RAIDs that go great together
Learn about the differences between RAID1+0 and RAID0+1.
- RAID shopping
It is best to test the leading vendors' equipment looking at the areas that you want to maintain availability, including the physical space and the personnel involved.
- How low will RAID costs go?
Affordable consumer storage networks may be right around the corner.
- These mistakes can kill RAID dead
There's a reason we still make backups of data stored in RAID arrays: Things can still go wrong that can cost you all or most of the data stored on the array.
- How to choose the right RAID for your array
Cut through vendor hype and determine which RAID level (or levels) is right for your disk array.
- Which is better: Cheap mirroring or RAID?
Want to avoid the cost and complexity of RAID? Mirroring is one option to consider.
- Do I have a RAID implementation?
Expert Brett Cooper gives the low down on what tools are available to see if a RAID is implemented on a server.
- RAID-2 isn't being used anymore
Even though you don't see RAID-2 being used all that much anymore, according to our expert Norbert Haag, and RAID-2 literature, there is a minimum number of drives when setting up a RAID-2 system. Hint: It's less than 10.
- RAID: How safe is your data?
RAID, in its current designs, can handle a single-disk failure gracefully. However, in the case of a double-disk failure in a given RAID volume the results can be disastrous.
- RAID groups and parity groups
RAID groups can contain single or multiple parity groups. You can think of the RAID group as the actual RAID container for data protection, and the parity group as a partition of that container.
- Migrating from DAS to a SAN
As for flexibility and performance, you can configure different RAID types within the arrays to match the applications requirements. RAID-1 or RAID-10 is best for log file storage, and RAID-5 can house database-type storage.
- Data compatibility between ATA and FC
ATA arrays may be configured the same as FC arrays (number of drives in the RAID group, stripe size and parity protection) or may be completely different, but that doesn't matter when the data is accessed from the storage controller by a server.
- Alternatives to software mirroring
The usual "ghosting" approach presents support challenges with RAID disks, making that solution unreliable in the current context. Mirroring does have its limitations, with the most obvious being the risk of corruption to two images instead of one.
- Checklist: Seven RAID configuration essentials
Here are seven quick tips for improving the performance and reliability of your RAID sets.
- Overhead for RAID-0+1 is too high … should I use RAID-5?
If you also think RAID-0+1 is too high, check out what expert Stephen Foskett has to say about using RAID-5 instead.
- Calculating available disk space in a RAID-5 set
Need to know how to calculate available disk space? Check out this Ask the Expert Q&A and find out how. Hint – you need to calculate usable space and the overhead of your RAID.
- Maximum number of drives in a RAID-5 set
The number of maximum number of drives in a RAID-5 set will vary by vendor implementation according to expert Greg Schulz. Take a peek at this Ask the Expert Q&A and find out more about drives in a RAID-5 set.
- Calculating space for parity in RAID-5
How do you determine how much space parity will consume in RAID-5? For instance, if I have a RAID-5 with 10 146 GB drives, what is the percent factor of space used for parity?
- Choosing the right RAID level
Find out what factors you need to consider when choosing the appropriate RAID level for a specific application.
- All RAID is not created equal
Although there is nothing wrong with using low-cost arrays, careful attention needs to be paid to the actual components of the arrays.
- How to select disk drives for DBMS
Database management systems (DBMS) put heavy demands on storage. If you have the opportunity to select disks specifically for a DBMS system, you can maximize your performance by making the right choices.
- Enterprise-class RAID functions: What they are and when to use them
This column explains these less familiar enterprise-class RAID features, starting with a set of features that ensure system uptime and flexibility.
- Advanced RAID functions
Part one introduced you to hot swapping, online capacity expansion and online RAID level migration. Part two exposes you to N-way mirroring, splitting and hiding and more.
- Expanding a four-disk RAID array
What do you do when your RAID 5 array is running out of room? Can you replace each drive one at a time and rebuild the array? A lot depends on which vendor you're using.
- In RAID-5, can parity information be exceeded?
Our storage expert Bruce Backa states a case as to why the RAID algorithm doesn't allow parity information to exceed its fixed-block size.