Mainstreaming SSD
Most of the major Solid State Disk (SSD) manufacturers and providers are reporting record sales both in terms of units and capacity being purchased. Much of this success is being driven by cost reductions in the technology and an increased understanding of how to best implement the technology. Mainstreaming SSD is going to require more than just price drops, its going to require intelligent leveraging of the technology.
Most of the major Solid State Disk (SSD) manufacturers and providers are reporting record sales both in terms of units and capacity being purchased. Much of this success is being driven by cost reductions in the technology and an increased understanding of how to best implement the technology. Mainstreaming SSD is going to require more than just price drops, its going to require intelligent leveraging of the technology.Most SSD implementations are pointed at a very specific application performance problem and the SSD tier is managed as its own very fast platform. The makers of SSD specific systems like Violin Memory, Texas Memory Systems and Fusion-io build systems or cards focused to ensure the surrounding system can deliver data as fast as the internal flash or DRAM memory inside can support it. As a result they sometimes don't have the data management features that users come to expect from the traditional SAN storage systems. As we discuss in our recent article "Integrating SSD and Maintaining Disaster Recovery", there are several ways to implement these systems to get the best of both worlds; high performance and high availability.
Broader implementation can come from system vendors integrating SSD into their systems. Companies like Pliant Technology and STEC have packaged SSD technology into mechanical drive form factors so storage manufacturers like EMC and others can integrate them into their existing drive shelves. Users storage virtualization technologies like DataCore or Falconstor don't even have to wait for their storage supplier, they can simply just build an SSD shelf and integrated it into the current storage framework.
NetApp has an interesting role in mainstreaming SSD. Not only can they virtualize an SSD system as evidenced by their relationship with Texas Memory but they also have their Performance Acceleration Module (PAM) which is essentially a flash based cache that installs inside a NetApp head to provided added performance without the special management steps of a stand alone SSD tier. Caching the front end of Network Attached Storage (NAS) has suddenly become a market of interest for many vendors.
Using SSD technology as a caching technology may be one of the key methods that mainstream SSD use. This allows a memory based cached (DRAM and/or Flash) to be used as a front end cache to traditional mechanical storage. Companies like Storspeed, Avere and Gear6 have developed technologies that help accelerate NAS based operations and companies like Dataram are providing similar technology for block based storage.
While each of these systems have different strengths they all allow users to get around the multi-tier management problem that SSD can cause. Equally interesting since these companies are leveraging an existing storage backend, with all its data management capabilities, they can focus on performance acceleration. SInce they may also be involved in every I/O that goes through to the storage subsystem, they also maybe able to provide very interesting analytics to users trying to understand where their performance bottlenecks are.
As the price and capacity of the SSD tier comes within the reach of an ever increasing number of data centers, access to this technology by a variety of methods becomes critical in making it a mainstream technology.
Track us on Twitter: http://twitter.com/storageswiss
Subscribe to our RSS feed.
George Crump is lead analyst of Storage Switzerland, an IT analyst firm focused on the storage and virtualization segments. Find Storage Switzerland's disclosure statement here.
About the Author
You May Also Like
Cybersecurity Day: How to Automate Security Analytics with AI and ML
Dec 17, 2024The Dirt on ROT Data
Dec 18, 2024