Performance

The PetaBlock is the ideal solution for organizations with significant archiving requirements. It offers all the convenience, security and data protection of the LTA solution, operating on the largest scale yet

  • Solutions starting at 240TB, scaling up to 5.7Petabyte in a single 42U rack. Multi-rack setups, also in geographically diverse locations, are supported
  • Archiving throughput of up to 32TB per day (over 3Gb/s: at such rates, ensuring sufficient network and source system capacity becomes an important part of implementation planning)

Media Choice

Advantages of the PetaBlock over other media types, such as Blu-ray disk or LTO tape include

  • Unparalleled data protection: using the recommended settings, up to 20 disks per availability group may be lost before data loss occurs
  • Media loss or degradation is detected quickly, and affected data is automatically moved to (or regenerated on) new media
  • Good concurrent archive/restore performance: data can be restored without impacting the archiving workflow
  • High throughput: media can be accessed without changer-related delays; multiple front-end servers can simultaneously access the same storage pool

RAID vs PetaBlock

Compared to traditional RAID solutions, the PetaBlock offers

  • Significantly reduced power consumption: smart scheduling ensures drives are only powered on when absolutely required
  • The ability to safely use cost-effective desktop-grade drives: since the duty cycle of each disk is kept to a minimum, no expensive ‘enterprise’ drives designed for 24/7 operation are required
  • The option to use drives from different vendors, of different capacities and firmware levels in the same storage pool
  • Effortless upgrades to higher-capacity drives using the LTA server’s migration functionality

Solution delivery

A complete PetaBlock system consists of the following elements

  • One or more LTA front-end servers, equipped with 12+ cores of processing power and 40TB of always-on RAID6 storage, responsible for gathering data and managing the PetaBlock infrastructure
  • Three or more 80-disk drive enclosures, running in JBOD mode, containing SATA disk drives with a capacity between 1 and 8TB
  • For systems with more than four drive enclosures or more than one front-end server: a SAS switching fabric is recommended

Petablock operation

The PetaBlock system archives your data using the following process:

  • The archiving front-end servers collect data from your network, using preset (scheduled) or ad-hoc archiving jobs. This data is processed and stored on the front-end server’s built-in RAID array.
  • At scheduled intervals, the pending archive data is transferred to the external drive enclosures. One enclosure at a time is powered on. In addition to writing the archive data, the front-end server will also, at this time:
    • Read any data required for any restore requests that have been queued
    • Perform housekeeping tasks, such as data integrity checks and re-generation of data lost due to drive failure or migration
  • The exact schedule is determined by your data volume, power budget, time-to-restore SLA and other operational requirements. For example, it’s possible to only gather data using non-peak hours (to reduce the load on the storage systems where the files to be archived reside).

Archive as a Service

The performance and scaling characteristics of the PetaBlock make it the ideal back-end for an “Archive as a Service” solution.

  • Multi-tenant, not just with unique encryption keys for each customer, but full metadata separation as well
  • External, third-party web site implements customer-specific functionality using the Archive Disk API, for example:
    • Authentication, integrated with existing user databases
    • Custom search engine to retrieve archive sets based on content or metadata
    • Per-retrieval billing and/or payment
    • Downloads of individual files from an archive set