Caringo has announced its second major software release in less than six months with the debut of CAStor 5.0, which adds named objects, multi-tenancy, dynamic caching, and remote data replication, as the company beefs up its object storage software platform for the private cloud storage market.
The updates found in CAStor 5.0 are all about semantics. Traditional file systems maintain symbolic names for folders and files in b-tree based directory structures, a practice that Caringo CEO Mark Goros claims limits scalability in both capacity and object count. CAStor 5.0 uses scaling to approach user-defined symbolic naming and provide access to billions of objects in a cluster. Goros said named objects semantics are consistent with the emerging de facto standards of public cloud storage services and work alongside CAStors system-assigned Universally Unique IDs (UUID).
It all starts with named objects. We used system-assigned UUIDs, which was a weakness for us in terms of the marketplace. Now users can use their own names alongside UUIDs for applications that require user-generate names, said Goros. We didnt plan on another major release this year, but we started working on new features that dramatically increased the number of use cases for CAStor and warranted another release.
The addition of named objects lead to multi-tenancy. CAStor 5.0 allows IT to host multiple departments or divisions within one physical private cloud , creating a common infrastructure for all applications, uses, customers and departments.
Within a domain, named objects are organized into multiple storage buckets that may be protected with their own security realms and Access Control Lists (ACLs). Users can also provide ACL control at the individual object level. New domains, buckets and objects are protected by CAStors automatic data replication functionality that replicates one or more copies to other nodes in the cluster to ensure continuous data availability, according to Goros.
In addition, CAStor 5.0 incorporates dynamic caching to accelerate read access of hot objects, similar to the way a Content Delivery Network (CDN) works. Heavily accessed objects are stored in RAM on nodes across the cluster to reduce latency and improve throughput. As demand for an object changes, CAStor automatically manages the cache to increase or decrease the number of cached copies throughout the storage cluster.
Caringo plans to make CAStor 5.0 available in mid-November.
Caringo released the previous version of CAStor last May with a power-saving feature called Darkive, which spins down disks and reduce CPU utilization based on predefined policies or predetermined levels of inactivity.
Darkive dovetails with CAStor's overall architecture, which is based on commodity hardware. The CAStor software turns off-the-shelf and legacy disk systems into storage nodes and dynamically balances load and storage based on usage and node specifications. Darkive can be applied to existing hardware with the only caveat being that legacy arrays contain disk drives with spin-down technology.
Individual CAStor nodes can be designated as an archival tier with disks and processors only spinning up when data needs to be recalled from the archive.
For related articles, visit Internet.com's new Cloud Computing site.
Follow Enterprise Storage Forum on Twitter.