The editors at SearchSOA regularly recognize particular services, middleware and architecture technologies for...
their innovation and market impact. The Clusterpoint 3.0 NoSQL database service is the July 2015 Innovation Spotlight selection.
Product: Clusterpoint 3.0
Release date: June 16, 2015
What it does
Clusterpoint 3.0 is a cloud-based database service that provides horizontal scaling for cloud environments. A distributed NoSQL database, the Clusterpoint technology is a multi-tenant system with a pay-per-use cost model.
"Having a single database in the cloud that can just subsume data and handle various workloads on demand eases a huge pain point of integrating several back ends," said Clusterpoint CEO Zigmars Rasscevskis, a former senior engineer at Google.
Clusterpoint 3.0 provides three key capabilities:
- The ability to scale instantly and execute queries on lots of servers in parallel without actually buying all of these servers
- A management console where developers can compare queries and immediately inspect results
- The ability to log all traffic against a database so developers can inspect to great detail the performance characteristics of different queries
Why it's cool
Clusterpoint 3.0 was created to tailor NoSQL for the cloud and bring cloud pricing to database services, two things other database vendors have not done, according to Rasscevskis. He finds this industry omission surprising because NoSQL's horizontal scaling is well-suited to cloud uses and can be expensive to deliver on premises. Only a few organizations have the funds to create massive hardware footprints to scale on premises, and this is why the cloud is the right place for NoSQL databases to run, said Rasscevskis.
Clusterpoint 3.0 enables a cloud in which only one NoSQL database is needed. For example, said Rasscevskis, a business typically uses a SQL database to manage transactional workloads and a NoSQL key-value store for the serving path and a search engine. Both contain the same data, which in turn requires developers to write integration code to keep everything in sync.
On the cost side, Clusterpoint 3.0 offers a pay-per-use model, disrupting the tiered pricing model of today’s traditional database-as-a-service (DBAAS) offerings, according to Rasscevskis. This pricing model differs from traditional cloud database pricing, in which customers pay for server instances.
Beta user Janis Rose, CEO of global direct mail provider Mailigen, chose Clusterpoint 3.0 to help customers handle fast-growing email volumes and pass on a per-use price.
Like many companies, Mailigen sees its and its customers' data volume growing rapidly and faces customers demand for real-time reports. "They need to know that we can support whatever they may need instantly and without the hefty price tag that comes with maintaining large amounts of data," said Rose.
Clusterpoint 3.0 fits the feature set of Mailigen's system, which supports various data workloads and types, including text, numeric and geospatial. One can dump data in a self-descriptive format -- XML or JSON -- and the database constructs all the right indices, Rose said.
The only downside is that a NoSQL database comes with its own query language, unlike SQL, which is standardized. Clusterpoint is no different, so moving from other databases to Clusterpoint involves adapting an application, said Rose. However, in practice, he's seen that that even moving from SQL, which involves changing the data model, can happen rapidly with Clusterpoint.
What is the price?
Clusterpoint's on-premises database software is free to use. A production system with a technical support package costs $300 per server per month.
For cloud deployments, Clusterpoint charges for resources used as follows:
- CPU core - 7 cents per hour
- Storage - 30 cents per GB a month, and technical support is included if resource cost goes above $200 per month
Discover Telerik for cross platform mobile development
Learn about MuleSoft