Embed
Email

How To Handle RAID Array Failures

Document Sample

Shared by: yaoyufang
Categories
Tags
Stats
views:
1
posted:
12/8/2011
language:
pages:
4
How To Handle Introduction

RAID Array

You implemented a RAID array to protect your data. While RAID

Failures arrays may fail less often than single drives, RAID array failure

does occur. Your organization should have many contingency

By Julius “Bud” Younke, CDRP,

plans in place for a variety of circumstances. What is your backup

MCSE, MCP + I, MCP, A+

plan when your RAID array fails?









Problem Statement





RAID1 and RAID5 (as well as others) are designed to provide high

availability of your data. RAID arrays that are truly redundant

(RAID0 is NOT a redundant configuration) are designed to

continue to function in the event that one drive physically fails.

They do not provide protection from logical corruption, file system

corruption or some physical issues such as bad, unreadable

sectors, or weak heads in the drive.



Any problems your RAID encounters which do not cause one of

the drives to completely fail and go offline can result in the RAID

volume itself going offline and not being accessible. While it still

exists on the drives, your data is effectively gone–locked away–

unable to be accessed by normal means.









Reclamere, Inc. • 905 Pennsylvania Avenue • Tyrone, PA 16686 • Phone: (814) 684-5505

How To Handle Issues

RAID Array

Failures (cont.) What RAID Level Is It?



By Julius “Bud” Younke, CDRP, Often a RAID volume that fails is a result of logical corruption in the

MCSE, MCP + I, MCP, A+ file system. Regardless of whether the issue is a file system fault

or an actual physical failure of the drive, more complex RAID types

often require professional recovery. RAID1 volumes are pretty

straightforward since they are simply the same data ‘mirrored’ onto

two drives. Only one of the drives needs to be recovered in order

to recover all your data. Attempting to recover a RAID array

without expert professional help can often result in unrecoverable

data.









How was the volume configured?



Striped raid volumes (RAID0 and RAID5) require a deeper

understanding of how the volume was configured. Drive order,

stripe size, parity direction, and offsets to the data--all items

defined by the particular RAID controller used–need to be

determined and accounted for in the recovery process.









What storage devices are connected to the network?



Storage devices connected to the network (NAS or SAN) can

represent their own issues. Simple consumer-type network

connected drives often contain two disks and boot a modified Linux

kernel as the OS for drive management. Part of the surface of one

of the disks will contain the Linux OS files to boot the storage









Reclamere, Inc. • 905 Pennsylvania Avenue • Tyrone, PA 16686 • Phone: (814) 684-5505

device, while the remaining portions of the disk are configured in a

How To Handle Linux soft RAID. This particular configuration is all but impossible

for consumer-type data recovery tools to recover.

RAID Array

Commercial NAS and SAN devices can also have some unusual

Failures (cont.) configurations. In a standard hard drive, a sector size is 512 bytes.

Some vendors use a sector size of 520 bytes on their SAN/NAS

By Julius “Bud” Younke, CDRP,

RAIDs. This results in recovery software looking for the 512 byte

MCSE, MCP + I, MCP, A+

sectors getting effectively ‘lost’ on the disk surface.









Should I attempt to recover the data myself?



While it is possible in some cases for a RAID volume to be

recovered using DIY means, it is much more complex than a

simple single drive recovery. As with any other type of recovery, it

is possible that simply trying to recover the data yourself can result

in more damage. Since RAID arrays often store mission critical

data or data subject to litigation, the best step when a RAID fails is

to contact a data recovery company with experience in RAID

volumes. Ask them about some of the RAIDs they have worked on

to see if they understand some of the intricacies of network

attached RAID devices. Be sure to ask if they perform ALL of the

work in-house or if they are simply a reseller of data recovery

services performed by others. If the data is subject to litigation or a

litigation hold, it is critical that the proper chain of custody be

maintained at all times.









Reclamere, Inc. • 905 Pennsylvania Avenue • Tyrone, PA 16686 • Phone: (814) 684-5505

Conclusion

The best protection for your data on ANY device is, of course, a

good backup policy with regularly tested backups. If a file is

deleted, or corrupted on the drive, no RAID can bring that file back.

RAIDs only purpose as far as redundancy is to maintain availability

of the data in the event of a disk failure. When it comes to

recovering your data when your RAID goes offline, an experienced

data recovery company is often your only choice.









Bud Younke, CDRP, MCSE, MCP+I,

MCP, A+



Lead Data Recovery Engineer and IT

Manager









Reclamere, Inc. • 905 Pennsylvania Avenue • Tyrone, PA 16686 • Phone: (814) 684-5505



Related docs
Other docs by yaoyufang
Catalog User Guide.doc - Firebrand Wiki
Views: 1  |  Downloads: 0
Slide 1 - University of California_ Berkeley
Views: 1  |  Downloads: 0
ASRF QUEENSLAND STATE COUNCIL
Views: 6  |  Downloads: 0
Web Design Final Project
Views: 0  |  Downloads: 0
Slide 1 - Law
Views: 0  |  Downloads: 0
CTC Job Search Outline
Views: 1  |  Downloads: 0
csepregi_kastely_angol
Views: 0  |  Downloads: 0
Table of Contents
Views: 2  |  Downloads: 0
By registering with docstoc.com you agree to our
privacy policy

You are almost ready to download!

You are almost ready to download!