I wrote (in French) about it a long time ago, in February 2007, when the product was introduced and I recently mentioned the product in the object storage timeline and history published in The Register July 15. In 2007, when the product was unveiled, HYDRAStor already offered Erasure Coding when some object storage vendors released it many years later in 2014 or 2015 claiming the real innovation. The product has already demonstrated its performance
- Scalability: this dimension is essentially a question of capacity. In a distributed environment, capacity is the sum of the individual storage space with a reduction and protection ratio.
- Resiliency: the idea here is to offer a system that can resist to errors but it also means durable data with some local or remote multi-copies or erasure coding techniques.
- Security: everything around access protection but also encryption and WORM capabilities.
- Performance: we don't speak here about super performance but good enough level to sustain a backup service, an archiving process or a tiering mechanism.
- Flexibility: we mean here essentially the access methods offered by the system to be integrated or connected to current environments with industry file sharing protocols such NFS or CIFS and of course Amazon S3 or OpenStack Swift but also Veritas OpenStorage Technology aka OST.
- Manageability: This includes the capability of the system to self manage, self tune, self repair and self scale when needed.
- Efficiency: essentially cost efficiency as storage efficiency is implicitly listed in the Scalability point. We mean here cost per usable TB and backup window satisfaction or more globally the quality of service delivered by the platform.
For these items, HYDRAStor offers native functionalities and among secondary storage solutions available on the market, this is clearly one of the most comprehensive product, no doubt.
|Scalability||165 nodes, 158PB, DataRedux (Inline Global Deduplication and Compression), Load Balancing of Data and Computing|
|Resiliency||Distributed Resilient Data (Erasure Coding), RepliGrid (WAN-Optimized replication, in-flight data encryption)|
|Security||Encryption-at-rest (AES-256), HYDRAlock (WORM), Regulations (SEC 17a-4), Dynamic Data Shredding|
|Performance||5.2PB/hour (inline backup speed)|
|Flexibility||File Sharing Protocols (NFS, CIFS), Object APIs (AMZN S3, OpenStack Swift), Veritas OpenStorage Technology (OST), 1GB or 10GB Ethernet, Multi-tenancy|
|Manageability||DynamicStor (Thin Provisioning), Self-Management, Non-disruptive addition/removal of nodes, Web GUI, CLI|
In addition of this features set, HYDRAstor has a strong ISV ecosystem for these IT services, facilitating the adoption. It will also help to consider the product as a consolidation of usages and replace independent, complex and costly data silos. NEC will participate to the next IT Press Tour in December and we'll learn more about HYDRAstor, wins, roadmap and key features.