Automatic Storage Tiering
Open vStorage has built-in auto-tiering functionality which automatically moves data which is cold to slower parts of the Open vStorage cluster. This allows frequently accessed data to be available in the fastest tier to ensure the the best performance and throughput.The multitude of levels is completely shielded away from the application and is transparently managed by Open vStorage.
Questions? We got answers!
Which types of storage media can Open vStorage use?
Open vStorage supports multi-level of tiering: RAM, PCIe flash, SSD, SATA drives and even amazon S3 compatible object stores or long term archives such as Amazon Glacier or Western Digitals Active Archive.
Why do I need automatic storage tiering?
Keeping all data all of the time on SSD will provide you with extremely fast storage. But the cost will be accordingly high. With Open vStorage it is possible to create an All Flash Array (AFA) by only using SSDs but Open vStorage allows to administrators to define automated storage tiering policies. For example, a data storage administrator can assign infrequently used data to slower, less-expensive SATA storage but allow it to be automatically moved to higher-performing SASor solid-state drives (SSDs) as it becomes more active (and vice versa). This allows to keep the TCO under control while delivering wicked fast storage performance.
Does Open vStorage support multiple tiers?
Yes. a typical Open vStorage cluster will use by default multiple caching and tiering layers. The most frequently accessed data will be stored in RAM, next come a global cache layer on SSD per datacenter and a distributed high-capacity tier across multiple datacenters. As final resting place for deep-frozen data Amazon S3 or another Open vStorage cluster can be used.
Can I use a public cloud as last tier?
Yes. Open vStorage is capable to use a public cloud such as Amazon S3, Microsoft Azure or Softlayer to store the cold data. Additionally you can spin up clones based upon the data store in the the public cloud directly in the public cloud.