Home  |   Customers    |   News & Events   |   Contact Us    |  Sitemap    
RAID Incorporated
Services Products Solutions Support Partners About
Case Studies
Request A Quote

Xanadu 120 Enterprise Modular Storage

Director Class Architecture

The Xanadu 120 is the industry’s only system to provide true modular expandability that allows for superior flexibility, performance and scalability in a director class solution.

Xanadu II

Overview

Xanadu II
  • Modular 5U Chassis / 2U Disk Enclosure design
  • 4 to 16 4Gbps Host Interfaces
  • 3 to 384 HDD
  • Dual Redundant Cache from 8 to 32 GB
  • SAS or SATA drives
  • High Speed X Bar Switch
  • Complete redundancy
  • Battery backup cache
  • SATA Guard Parity Validation
  • Self Healing
  • MAID Technology
  • Dynamic Provisioning
  • Full enterprise software suite available

Scalable Capacity and Performance

  • The Xanadu 120 single node supports up to 384 drives per module.
  • Customers who start with a Xanadu but later discover they need more capacity than the Xanadu allows, can convert the Xanadu 110 array to Xanadu 120 without losing or migrating data.
  • Xanadu 120 will be available in single, dual or quad configurations supporting up to 1,536 SAS or SATA disk drives.
  • Xanadu 120 can be upgraded in real-time from single to dual or quad, non-disruptively, so no data or uptime is ever lost.
  • The high-speed X-bar switch architecture enables true multi-node operation, with performance scaling linearly with capacity.
  • All models are managed through a common software package called StorageManager.

Xanadu Upgrade Path

Director Class Architecture

The director class architecture has the flexibility to expand each of it’s components where they are needed within each system.  Each individual 5U system consists of a Host Director, a Cache Module, a Disk Director, an X-Bar switch.  Disk drives can be expanded in 2U twelve drive increments. Each 5U system can be expanded up to a total of 4.

Add components individually,in real time, non-disruptively!

Director Class Architecture

X-Bar Architecture True Multi-node Operation

Multi-node Operation
  • The high-speed X-bar switch architecture enables true multi-node operation, with performance scaling linearly with capacity.
  • Each system can be  interconnected with up to 3 other systems by utilizing patented X-Bar  technology.
  • This results in a maximum scaled configuration of 128GB of cache 64 Fibre Channel ports and 1536 HDDs.

SATA Guard

  • Recent large academic studies have identified the surprising frequency of silent read failures.
  • Such errors result in corrupt data being provided by the disk array to the application without any notification.
  • This “silent data corruption” includes misdirected writes, partial writes, and data path. 
  • The Extended Data Integrity Feature detects and addresses silent data corruption, while also preventing parity pollution.

Failure Avoidance: Self Healing

Failure Avoidance

  • In traditional disk subsystem designs, sector errors would mark the disk drive as failed, initiating a RAID rebuild process.  This process degrades performance and takes a long time to finish. It is also expensive as there may still be useful life in the disk drive. 
  • Patented Phoenix Technology detects these errors but allows operation to continue with the other drives in the RAID group. If an alternative sector can be assigned, the drive is allowed to return to operation with the RAID group without having to do a complete rebuild.
  • Phoenix technology allows performance to be maintained throughout the detection and repair process.

Dynamic Provisioning

Dynamic Provisioning

Traditionally, logical disks could only be managed in units of the physical sets of disks that made up the RAID. Users had to install a fixed number of disks even if they were not needed. The new dynamic provisioning function permits logical disk capacities to be changed dynamically to either expand or shrink. Physical disks can  be added one at a time. The logical disks required can be managed efficiently, and the system is free from physical restrictions.

Features:

  • No data loss up to two disk failures
  • Two modes: 4+PQ or 8+PQ
  • LUNs spread across entire pool in data stripes similar to RAID-6
  • Dynamically add HDD one by one
  • Dynamically grow/shrink LUNs in pool
  • Better performance than basic pool RAID-6

Dynamic Provisioning is a unique storage management and protection implementation which provides the power of RAID-6 data protection with the flexibility to manage capacity one drive at a time. RAID-6 has a higher capacity overhead than RAID-5, but can protect against data loss from two disk failures, which RAID-5 cannot.

  • In a Xanadu 120 Dynamically Provisioned pool, the number of drives participating in a RAID-6 pool is not fixed and can be dynamically expanded as capacity requirements dictate.
  • Traditional RAID implementations can be expand only in full multiples of the RAID group size.
  • Dynamic Provisioning does this by maintaining RAID-6-like parity protection in data stripes rather than full disk drives. When a logical disk is defined, it spreads the stripes across all drives in the pool.  Disk drives can be added to the pool at any time, one at a time, without reformatting the raid group or moving any data files.  When a new disk drive is dynamically added to the pool, some of the strip stripes from each of the logical disks are moved to the new disk drive. The next time a logical unit is allocated, it will be spread across all drives in the dynamic pool. 
  • Performance in a dynamic pool should be better than in traditional, fixed-sized pools because all the drives in the pool are available to service I/O requests, not just the number of drives in a traditional RAID group.

MAID: PowerConserver

Massive Arrays of Idle Disks: Power down inactive disks – lower power and cooling

MAID: power conserver 30-60% energy savings in secondary storage applications

High Availability:  PathManager

Xanadu 120’s PathManager software provides a convenient combination of path failover and load balancing across multiple Fibre Channel paths between the host and the array. Should one path fail, all data traffic will be rerouted over the alternate path to continue operation uninterrupted. PathManager provides two types of load balancing. Static load balance directs traffic to specific logical disks to follow specific paths. Dynamic load balancing spreads traffic to all logical disks across all available paths. Path Patrol periodically checks paths the ensure that they are still available, and also periodically checks failed paths to detect when they have been restored.

Pathmanager

The convergence of modular and enterprise features on a single platform, the scalability, availability, and manageability that provides, now gives customers the best TCO and ROI available today.

Call for Optional MAID and WORM Features

If you would like pricing or additional information, please fill out our Request a Quote form, call us anytime at 800-330-7335 or email sales@raidinc.com.

PRODUCTS

OVERVIEW

HPC

Xanadu 240

Xanadu 230

Xanadu 220

X2-IB

HPC Storage Solution

Fusion Gateway

PRODUCTION

XANADU 130

XANADU 120

XANADU 110

XANADU 110 iSCSI

ARCHIVE

FALCON IV

FALCON III (48-BAY)

FALCON III

MAVERICK II

METADATA

RAZOR

RAZOR SSD

 

Whitepaper:
Silent Data Corruption in SATA Arrays

Download here>>

Whitepaper

RESOURCES

Downloads
Case Studies
Why RAID?

  Home   |   Customers   |   News & Events   |   Request a Quote   |  Contact Us
Services   |   Products   |   Solutions   |  Support   |   Partners   |   Company
Employment   |  Legal   |  Join Our Mailing List   |  Directions   |  Privacy   |   Sitemap
© Copyright 1994-2010
RAID Incorporated™
Sales/Support: 1-800-330-7335