People often ask what sets SimpliVity apart from other hyperconverged infrastructure vendors. We have many answers to this question, but underlying them all is the data efficiency built into the OmniStack Data Virtualization Platform. Earlier in the year, we kicked off a series to discuss our data efficiency technology and the business advantages that customers can discover as a result of them. So what should you take from this series?
The rapid increase in the amount of data, coupled with the stagnant increase in performance (with the notable exception of flash-based storage devices) have led to a painful situation in the data center. This situation is a collection of problems: noisy neighbor problems, missed recovery point objectives (RPOs), shrinking backup windows (with increased amount of data to protect), recovery time objectives (RTOs) that don’t meet business demands, multiple storage silos containing the same data, and wasted IOPS due to unnecessary copying of data during tiering, backups, caching, etc.
Originally designed to help solve the pain of uncontrolled growth of data and the storage capacity needed to capture it all, data efficiency features (including compression, thin provisioning, data deduplication) are table stakes in today’s storage arrays. This is also the basis of other data management appliances in the data center (backup, WAN optimization, archival systems, all-flash arrays). Unfortunately, these efficiencies are generally only focused on managing the capacity problem, not the underlying issues actually causing these pain points, which are all fundamentally performance and management problems.
SimpliVity set out to solve all of these problems, focusing not only on the primary data storage, but across the entire lifecycle of the data. The result was the OmniStack Data Virtualization Platform, which improves the performance and capacity optimization of our customers’ virtualized server environments across all their sites, including multiple data centers, remote and branch offices, and disaster recovery sites, whether on-premises or cloud-based.
In order to perform these data efficiency technologies most effectively, they must be performed inline as the data is ingested into the storage subsystem. Deduplication and compression, coupled with a focus of managing data across all the phases data goes through in it’s lifecycle, were utilized in the SimpliVity architecture primarily to help reduce the number IOPS being sent to the hard disk drives. The impressive capacity savings is a nice side effect for SimpliVity customers.
Unfortunately, both deduplication and compression are compute-intensive operations; therefore, they require a significant amount of CPU resources. Not a great model for a hyperconvergence platform that also needs the CPUs to serve the virtual machines. In order to provide all these services in a hyperconverged infrastructure, SimpliVity delivers our software with hardware-assisted deduplication, courtesy of the OmniStack Accelerator Card. This allows the OmniStack Virtual Controller to free up CPU resources (and licensing costs) on the host for use by the VMs, while also providing a very predictable, high performance, low latency hyperconverged platform.
This focus of data efficiency across all phases of the lifecycle means that data does not need to be taken out of the efficient state to be backed up, moved across the WAN or stored in cache. By extending this platform across multiple sites (including cloud storage platforms like Amazon Web Server) and tracking data deduplication in a globally-aware fashion, remote data operations also become more efficient, improving WAN utilization and the speed of data movement. By maintaining deduplication and compression within cache, more data can be stored in RAM or SSD. Since VMs are stored and backed up as logical constructs, we are able to clone, backup or restore a local VM in less than one minute with little to no IO.
This has been the core to the SimpliVity hyperconverged platform since the foundation of the company. Our customers have seen these benefits consistently since we shipped our first product. So consistently, in fact, that we now are putting a guarantee directly into standard warranty. How can you not want to try it!?
To delve deeper into these advantages and the underlying technology that makes it possible, I encourage you to review our Data Virtualization Platform Whitepaper.