Contact Sales

The Scalability-Reliability Dilemma

Hard drive failures are inevitable, regardless of the storage platform. With scale, drive failures become more frequent simply because there are more drives to fail. For this reason, most scalable NAS systems present a dilemma: scalability that comes at the cost of lower data reliability and availability.

After drive failures, all data is at higher risk as the storage system runs in a degraded rebuild mode until the rebuild completes. As a result, long RAID rebuild times lower overall data availability. The quantity of data to rebuild and the performance of the RAID architecture determine the rebuild time. Traditional hardware RAID solutions were architected for protection, but not fast rebuild performance. This introduces greater risk to production data for the hours or days it takes a rebuild to complete, particularly when storage is deployed at scale.

Click Image to enlarge

PanFS® RAID 6+ Revolutionizes Reliability and Availability

With PanFS, data reliability and availability on an ActiveStor solution actually increase with scale, not the other way around. Its innovative per-file RAID architecture with RAID 6+ triple-parity data protection actually lowers the percentage of files affected by too many drive failures as the system scales. This establishes a new standard for enterprise-grade data reliability.

PanFS RAID 6+ is far more robust than a typical RAID 6 implementation, providing availability and reliability at scale that meets the requirements of business critical workflows. Per-file RAID with triple parity protection work together to minimize the amount of data that must be rebuilt after one or more failures. Finally, PanFS is a boon to business continuity efforts as it is able to maintain availability to all undamaged files even after three drive failures where other storage systems would have gone completely off-line. A limited number of specific files can normally then be easily and quickly restored from backup or a replica file system without the need for a multi-petabyte restore that would otherwise take days or weeks to accomplish.

 Data ReliabilityVolume of Rebuild DataTime to Rebuild DataExtended File System Availability (EFSA)
Legacy NAS w/ RAID 6
  • Decreases with scale
  • Substantial amount of data to be rebuilt after disk failure
  • Rebuilds damaged and undamaged files
  • Higher risk of prolonged exposure in degraded mode
  • Rebuild in days
  • Rebuilds performance not scalable
  • Serial data channels slow rebuilds
  • File system not available after triple disk failure
  • Complete file system restore often required
PanFS w/ RAID 6+
  • Increases with scale
  • Minimal data to be rebuilt after disk failure
  • Only repairable files rebuilt in most cases
  • Lower risk of prolonged exposure in degraded mode
  • Rebuild in hours
  • Rebuilds performance scales with the system
  • Parallel data channels for faster rebuilds
  • File system is available after triple disk failure
  • Restores limited to a small list of specific files in most cases

Business Benefits of PanFS RAID 6+

The business benefits of RAID 6 dual parity protection have been established for years. The benefits of PanFS RAID 6+ triple parity protection expand on the proven data protection benefits of RAID 6.

The enormous data growth rates seen in technical research and enterprise environments have made data loss more likely than ever before. Even with protection against double drive failures, data remains at high risk during long disk rebuilds that often take days to complete on other storage systems. While RAID 6 greatly minimizes the chance of data loss relative to RAID 5 protection, the business consequences when a loss occurs can be extremely disrupting to production. These can range from low performance and associated workflow slowdowns to complete unavailability of file system data — bringing production to a complete stop as restore from backup takes days to complete.

Many environments require enterprise data reliability as any significant disruption to production, discovery, and innovation cannot be tolerated — even once. PanFS RAID 6+ sets a new standard of enterprise availability for these environments to ensure ongoing productivity.

Extended File System Availability (EFSA)

PanFS RAID 6+ protects file system directory data one degree deeper than file data. This allows the file system to remain available even after a triple disk failure. Although a small percentage of files may be unavailable, the file system will normally remain online, allowing unaffected files can be accessed normally. During this period of extended availability, full data reliability and availability can typically be regained by restoring the specific files that were affected from backup. Without Extended File System Availability, IT may be faced with performing a complete restore of a multi-petabyte file system.