Specifications

ES/OL
Evaluator Series On-Line™
Storage Systems
Product Analysis
May, 2011
Page 14 of 30
Analysis of Fujitsu ETERNUS DX400 S2
Models DX410 and DX440
© 2007-2011 Evaluator Group, Inc. All rights reserved. Reproduction of this publication in any form without prior written permission is prohibited.
7720 E. Belleview Avenue • Suite 210 • Greenwood Village, CO 80111 • [phone] • Fax [phone] • www.evaluatorgroup.com
RAS FEATURES
ETERNUS storage systems maintain data integrity and provide data availability via the following mechanisms:
• RAID protection of disk devices
• Redundancy of active components
• Dual power source
• Write cache data de-stage to disk
• Concurrent configuration changes
• Hot spare disk drives
• Concurrent maintenance on components
• Concurrent firmware upgrades
• Parity protection
• ECC (Error Correction Code)
• BlockGUARD
Data Integrity
Data integrity is expected of any storage system and the methods that system vendors employ to ensure data
integrity are important for any analysis. Fujitsu has implemented an end-to-end integrity checking mechanism
that is called BlockGUARD. BlockGUARD puts an extra eight bytes of information on every 512-byte data
block as it enters the system at the channel adapter. These eight bytes are retained with the data as long it is
stored within the system. The eight bytes are composed of a four byte cyclical redundancy check (CRC) and
four bytes that are the block ID. This information is checked whenever the block is stored on the disk by the
device adapter and is also checked when data is retrieved, both at the device adapter and the channel
adapter. The BlockGUARD characters are stripped whenever data is sent to the host by the channel adapter.
Data integrity is also a part of the cache function. Data in cache is error-correction code (ECC) protected to be
able to detect and correct cache memory errors.
As described previously, Fujitsu also supports Oracle HARD for end-to-end data protection with Oracle
databases.
Redundant Copy
When a disk is identified for preventive replacement (before hard failure), the redundant copy function builds
the identified disk’s data from the other disks in the RAID group to a hot spare. Once that action is complete,
the disk at risk is detached and the hot spare is incorporated into the RAID group ensuring data integrity. This
is accomplished by using S.M.A.R.T (Self-Monitoring Analysis and Reporting Technology) technology in the
disk drive to collect patrol and error statistics.
Redundant Power System
The power system is based on redundant power supplies in the modular enclosures with dual line cord inputs.
The modular architecture and dual power inputs ensure the greatest amount of power problem isolation. All
power system components are replaceable non-disruptively.