This big-data startup wants to manage your data lake in the cloud
- 15 December, 2015 23:04
Big data may offer companies a world of untold potential, but realizing the benefits is typically no walk in the park. That's why "big data as a service" platforms have begun to emerge, and it's also why Bigstep is taking a like-minded approach to the increasingly common data lake.
Launched last year, U.K.-based Bigstep has been offering big data as a service through its Full Metal Cloud platform, which already includes compute instances, block storage and network components. Now, the Full Metal Data Lake extends that platform to include exabyte-scale storage for big-data workloads as well.
“Businesses today have access to infinite amounts of data but no fast, easy or cost-effective way to make sense of it,” said Flaviu Radulescu, Bigstep's CEO.
The Full Metal Data Lake aims to help companies get actionable insights from their data in just a few clicks, Radulescu said.
The Full Metal Data Lake can be used in stand-alone mode for active data storage, or it can be integrated with on-premises infrastructure; it can also be used with any application or service in the Full Metal Cloud platform. The service can be activated instantly, and requires no configuration or minimum commitment, the company says.
All files are encrypted both in transit and at rest; data can be imported or exported via encrypted tunnels or SSL connections. The platform can accommodate both structured and unstructured data, either continuously streamed or imported in batches.
The Full Metal Data Lake is compatible with the Hadoop Distributed File System and can be integrated with tools including not just Hadoop but also Apache Spark, most NoSQL databases -- including Couchbase, Cassandra and Redis -- and products such as Elasticsearch, Solr, Qlik, Tableau and R. It works out of the box with container technologies such as Docker, Mesos and Kubernetes as well.
The new data lake as a service is priced at £18 per month for each terabyte of data, with no additional upload fees. Bigstep has offices in the U.K. and Bucharest, with infrastructure in the U.K. and Germany.
Research firm MarketsandMarkets expects the big-data-as-a-service market to grow from $1.8 billion this year to $7 billion in 2020.
Big data as a service offers a level of simplification and abstraction "away from the underlying technical complexities that even Hadoop in the cloud-type offerings from AWS or Microsoft offers," said Brian Hopkins, a vice president with Forrester Research.
Qubole, for example, is one provider that offers a service tier that it manages in the cloud.
"You load your data into AWS's stores like S3 or Redshift," Hopkins explained. "They provide the connectors to that data from their service tier. Then you access the data via MapReduce, Spark, Hive, etc., through their service."
Such vendors simplify the scaling and management of service levels, security and metadata-management aspects, he added.
Cazena is another contender, and it offers a data-lake-as-a-service product of its own, Hopkins pointed out.
Roughly 36 percent of enterprises are considering cloud services for Hadoop, and 41 percent are mulling them for Spark, said Nik Rouda, a senior analyst with Enterprise Strategy Group, citing research his company has done.
Bigstep isn't the first vendor to offer cloud-based big data, but it could be set apart by its strengths in security, managed service and setting up hybrid environments, he said.