HIOPS Compression™ represents the next generation of compression technology from Permabit. HIOPS delivers advanced, high performance, inline compression that is uniquely optimized to support OLTP and Big Data workloads in primary storage environments. Unlike most compression solutions on the market, which run on a single CPU core, HIOPS leverages multi-core processors to run parallelized packing algorithms that handle many compression operations at once, enabling it to deliver the fastest compression in the storage industry.
HIOPS Compression is offered as a separately licensable component of Permabit’s Albireo VDO (Virtual Data Optimizer) product. Albireo VDO is the first and only data reduction software to offer inline data deduplication, compression, and thin provisioning in a single solution. All three technologies (compression, deduplication, and thin provisioning) are must-haves for storage vendors that wish to address typical F1000 performance requirements and to differentiate their products based on the availability of superior data reduction technologies.
Key Features and Benefits
|Data Compression||Reduces the effective cost of storage by as much as 5X without performance impact. Delivers increased storage capacity across structured and unstructured data sets.|
|Multi-core scaling||Leverages today’s multi-core processors with massive parallelization to deliver industry leading 650K IOPS (4 KB) in mixed (70/30) read-write operations. Supports performance required for OLTP operations.|
|Flash, Disk, and Hybrid Optimized||Delivers consistent, inline performance.|
|Deduplication integration||Works in concert with compression as part of Albireo VDO, offering the only data reduction solution to provide inline high performance for BOTH deduplication and compression.|
How it works
HIOPS leverages Permabit’s advanced deduplication capabilities and allows VDO inline deduplication to be applied first. This eliminates duplicate data immediately and reduces the need for unnecessary compression steps, saving CPU cycles. After initial deduplication, any unique data remaining is compressed. The HIOPS “dedupe first” approach stands in contrast to other solutions on the market that compress data before doing deduplication (due to the performance overhead of their deduplication operation), which requires additional CPU cycles and reduces overall performance. HIOPS uses a fast, parallelized packing algorithm to find multiple compressed blocks that can be written together within a single physical block. This saves processing cycles, reduces storage consumption, and increases compression throughput efficiency.