Long-time archiving PetaBlock

The Archiveware PetaBlock is a high-performance, yet low-cost appliance for long-time archiving. Using high-density enclosures, containing 90 disk drives each, it can be used to routinely transfer 5 terabytes of data every day. If additional throughput is required, the PetaBlock can be equipped with an even higher-performance server, provided sufficiently fast source storage systems as well as a 10 Gigabit network are available.

Long-term availability of archive sets is ensured using data redundancy. With erasure coding algorithms and overhead similar to traditional RAID, up to 18 drives may malfunction simultaneously in a given enclosure. This extraordinary resilience, which can even be increased if desired, not only ensures peace of mind, but also allows failed drives to be replaced on a fixed schedule (for example once per month), instead of as a matter of urgency. For ultimate data protection, the contents of a primary PetaBlock can be replicated to a backup location. Energy consumption and cooling requirements, and thereby total cost of ownership, are kept to a minimum, by only powering enclosures when actually required for data transfer or maintenance operations. Depending on the total system capacity and how often restores are required, this can reduce electricity expenses by up to 90%, while still allowing immediate access to data if necessary.

As part of the PetaBlock maintenance processes, stored data is regularly checked for consistency. Each archive set section contains identification details, including a checksum, allowing it to be replaced immediately in case it is damaged or missing. In contrast to archiving solutions based on tape or Blu-ray, any storage media degradation is detected and resolved automatically.

Moreover, unlike traditional RAID arrays, the PetaBlock does not require matching disk drives, allowing media from multiple vendors and of different types and firmware levels to co-exist in a single enclosure. This lets you use the most cost-effective media, as well as painlessly migrate away from drive models that turn out to be less reliable. Cross-drive and cross-enclosure migration is automatically managed by the Archiveware software, making it easy to decommission old enclosures upon adding new higher-capacity ones.

Archive data can be exchanged using standard file sharing protocols (CIFS/NFS) as well as a S3-compatible REST API. For workflow and data lifecycle management applications requiring additional control over the archiving process, Archiveware offers custom integration services. Data at rest is fully protected using AES encryption, with key management handled by a built-in Public Key Infrastructure, optimized to ensure long-term key material availability. Metadata added at the time of data ingestion is used to enable queries of archive set content and to select sets to be restored.

Metadata can also be used to assign retention and protection policies. For example, certain archive sets can be automatically deleted after some time, whereas others are transferred to another archive tier, such as Blu-ray-based cold storage.