Engineering a World-Class, Big Data as a Service Solution for a Top Five Cloud Services Provider

Categories: Partners Success Stories

The Challenge: Driving Differentiation for Cloud-Services Providers

Cloud providers are facing increasing competition for market share.  As a result, many are looking to value-added services to differentiate their offerings in high-demand areas such as development operations, mobile, and big data.  These offerings provide customers with out-of-the-box solutions designed to accelerate new projects, and they encourage customers to concentrate on business solutions rather than infrastructure design, implementation, and support.

SoftServe’s customer, a top five cloud-services provider, wanted to differentiate their offering through a Big Data as a Service solution to and asked us SoftServe to design and implement a complete solution to help serve their customers.

Big Data as a Service

Through talking with the customer, it became clear that their needs included complete automation, self-recovery from deployment failures, auto-scaling, self-healing, capacity planning, as well as usage and billing analysis capabilities. In order to meet those requirements, we needed Apache Hadoop to integrate and analyze all the data.

In cooperation with the customer, we conducted a comprehensive analysis of existing Apache Hadoop distributions, which showed that Cloudera offered the best toolset on the market to make implementation of such a challenging project achievable.  Cloudera was simply faster, easier to manage, and more secure. We deployed Cloudera Enterprise, starting from a comprehensive and rock-solid backend, and ending with a flexible and multi-platform UI/UX for a self-service portal, which can serve thousands of customers and provide them with high quality big data cloud services.

Cloudera’s tools – including Cloudera Director, Cloudera Manager, and their state of the art API and flexible configuration capabilities – played a major role in provisioning and bootstrapping cluster nodes and components in just minutes. Additionally, a set of Ansible scripts were developed to automate cluster post-configuration and support failure-recovery, self-healing and automatic scaling. This lightweight, but powerful self-service portal was developed using the most advanced WebUI technologies like HTML5, CSS3 and JavaScript/Ajax.

The Success!

From a business perspective, this application has established a competitive edge in the world of cloud-based Big Data solutions and has become a significant part of the customer’s cloud portfolio. Unique integration of Cloudera Manager and Cloudera Director, without complex intermediate software allows customers to use the power of Apache Hadoop  to maximize value for their businesses, while services like auto-scaling and failure-recovering allow better resilience and response to the customer’s load.

Most importantly, the Big Data as a Service solution developed has demonstrated immediate efficiency and resulted in significant new Fortune 100 customers… not to mention praise from the customer’s sales team for building a new cloud service, which allows them to close big deals!
Valentyn Kropov is a Solutions Architect at SoftServe, a Cloudera partner specializing in the design, implementation, performance tuning, and management of Big Data solutions.


Leave a Reply