Permabit Enterprise Archive Architecture
Massively Scalable Storage Grid Architecture
The Permabit Enterprise Archive™ architecture is based on a storage grid design, comprised of Storage Modules, Access Modules, and network switches. Storage Modules and Access Modules scale independently to enable more performance on input (Access Modules) or if additional storage capacity is needed individual Storage Modules can be added. This flexibility delivers easy, cost–effective expansion to meet future performance and storage needs.
Access Modules
Access Modules are the interface between the network and the Storage Modules. Access Modules perform digital fingerprinting checking to determine if the data is already in the storage Modules and Scalable Data Reduction™ (SDR) provides in–line sub–file level data deduplication with compression to dramatically reduce the total amount of storage needed to store data.
Access nodes are paired for high availability; ensuring end users always have access to their data. Rather than waiting for a failover with one node active and the other passive, both nodes share the workload. One node handles incoming and outgoing archive transactions, while the second maintains an active journal of those transactions. Should the primary node fail, the secondary node assumes control and continues writing information seamlessly without user intervention and more importantly without data loss.
Storage Modules
Data routing and indexing occur in the Storage Modules. Each 2U Storage Module is comprised of two independent hot swappable nodes, each of which contains 6TB’s of storage providing a combined capacity of 12TB’s per Storage Module. A typical 42U rack will house 2 Access Modules and up to 18 Storage Modules in a storage grid. All storage within the grid is virtualized, resulting in zero provisioning and effortless LUN management. As additional Storage Modules are added to the grid, the new capacity is automatically incorporated into the storage pool allowing seamless storage growth without user interruption. This feature also aides in decommissioning older hardware. Simply remove a storage node and the system automatically redistributes the data to other nodes and notifies you when to remove the old node. New technology can be added and older technology can be migrated out without ever having to perform a fork–lift upgrade and time consuming data migration.
Data is distributed across Storage nodes using RAIN–EC technology. Patented RAIN–EC technology enables the Permabit Enterprise Archive to scale to multiple petabytes and delivers the most reliable storage technology available in the industry. With RAIN–EC, a storage grid can survive multiple simultaneous component failures without compromising data and recovery from node failures is much faster than RAID technology because all drives in all nodes participate in the recovery process. Customers typically forego the expense and hassle of traditional backup and instead couple Enterprise Archive with a Replicated node for disaster protection.
Data Migration
Figure 1.
Applications can write directly to the Permabit Archive using standard CIFS and NFS interfaces without requiring costly API’s. Alternatively, third party data movers, from vendors such as Atempo, CommVault, Symantec or CaminoSoft, are available and these products can also write directly to the archive using the standard CIFS and NFS interfaces. The Permabit archive can be used to answer a wide range of general purpose storage requirements.
Permabit Enterprise Archive Architecture Features:
- Storage Grid: Allows for independent scalability of Access Modules and Storage Modules.
- Access Modules: Provide access via standard NFS and CIFS interfaces to Storage Modules. Can be added independently of Storage Modules to increase performance.
- Storage Modules: Provide massively scalable, reliable storage. Can be added independently of Access Modules to increase storage capacity.
- Massively Scalable: Can have up to 18 storage node pairs per grid for a total of 216 TB of raw capacity.
- Intelligent Recovery™: RAIN–EC is designed for scalable data protection in multiple petabyte archives.
- Zero Provisioning: No more complex LUN management. Full storage management automation.
- Mix and Match: Today’s storage technology can work with future storage technology without ever having to perform a forklift upgrade.
Key Benefits
- Stores massive amounts of data in a fraction of their original storage footprint, driving effective cost per gigabyte to less than $1.
- Transparent expansion simplifies migration and eliminates expensive future upgrades.
- Effortless storage management removes costly provisioning.
- RAIN–EC delivers scalable data protection and removes need for traditional backup.
- Survives multiple simultaneous failures (multiple drives, power, CPU, memory, etc.) without data loss.
- Reduces data center storage footprint by storing more data in less physical space.