Focused squarely on private cloud storage, privately-funded startup ParaScale provides its ParaScale Cloud Storage (PCS), a software-only solution that enables hundreds of commodity servers to be clustered together to act as a file repository with massive capacity and parallel throughput. This object-oriented solution is loosely coupled, where the nodes don’t need to talk to each other much. Being loosely coupled allows huge scalability and parallel performance for multiple files (or multiple copies of a single file) across multiple nodes. Other proprietary examples include the Google file system and Amazon S3, and open source technologies like Hadoop.
By way of contrast, tightly clustered file systems (Bluearc, Pananas, NetApp Spinnaker/ONTAP8, iBrix, HP/PolyServe, etc.) are tightly coupled and were built to solve the challenges of high performance computing.
ParaScale's Policy Engine is a rules engine that runs on a control node. The Policy Engine simplifies management by synchronizing and administering hundreds of servers and thousands of disks. The policies managed by the engine continually tune the performance of the storage cloud to achieve the desired objectives like maximizing disk availability and read bandwidth, and minimizing access latency.
Thus, PCS is extremely flexible. It can be configured for video streaming, video editing, parallel data capture, massive read-parallelism applications such as HPC or data warehousing, and massive write-parallelism applications such as video surveillance.
Note, that PCS is not positioned as tier 1 storage as it is missing a few availability features, but we expect this to change within a year.
Technology Highlights
- Supports standard protocols (NFS,HTTP,FTP,WebDEV)
- Global namespace (can be partitioned)
- Control nodes (2 per cloud) and data nodes (currently 100 max.)
- Thin provisioning
- Write once, consume many
- Massive read bandwidth
- Parallel streaming writes
- Automated load balancing and sharing
- Automated capacity balancing
- Policy-based replication/data redundancy (a key enabler)
- Scale up and/or scale out
- Self healing
- Access Control Lists (ACLs) and file read/write locking for security and data consistency
- Runs on Linux and uses EXT3 to actually store the files
- User applications can reside on same node
Customer Highlights
- Beta trials were mixed about 50/50 between services providers and enterprises building private storage clouds. Both models appear to be gaining traction.
- PCS is being used by enterprise and service provider organizations including:
- Blue Coat Systems
- Carpathia Hosting
- Sony Pictures Imageworks
- Stanford Genome Technology Center.
- Technology providers including:
- Vembu Technologies
- South River Technologies
- ISP in the Midwest for secure tier 2 NAS.
Company Highlights
- Team
- Deep clustering, storage expertise (66 patents amongst the team)
- Team from Veritas, HP, Sun, Locus, Teradata, Scale8 and NetApp
- History
- Founded 2004,
- Alphas running software since early 2006
- Beta since mid-September 2008
- GA March 2009(?)
- Investors / Funding
- Menlo Ventures and Charles River Ventures
- Successful investment track records within clustered storage (Spinnaker / EqualLogic)
- $11.37 Million to date, Series A - May 2008
- Website
Pricing
- List pricing is a one-time charge of $US 1.05 per gigabyte
- First four terabytes are free
Action Item: For users who want scalable storage for less money, ParaScale definitely warrants consideration. The most cost-effective way to use ParaScale is to take advantage of the internal storage on Linux servers, but customers can also connect to Fibre Channel, network-attached storage and other types of storage.
Footnotes: