Skip to content

Data Centre Operations to Run on Hadoop According to WANdisco's Announcement

In the year 2000, Dr. Yeturu Aahlad, then leading the distributed computing department at Sun Microsystems, decided to leave his position to tackle a challenging computer science puzzle that piqued his interest.

Data Centre Operations Set to Switch to Hadoop, According to WANdisco's Statement
Data Centre Operations Set to Switch to Hadoop, According to WANdisco's Statement

Data Centre Operations to Run on Hadoop According to WANdisco's Announcement

WANdisco's Non-stop NameNode Revolutionises Big Data

In the realm of big data, Hadoop is poised to become the operating system for data centers, replacing Linux, according to David Richards, co-founder of WANdisco. This prediction is backed by the success of companies like WANdisco, which has won customers such as HP, Intel, and Lockheed Martin.

WANdisco, founded in 2005 by Aahlad and UK IT entrepreneur David Richards, has made significant strides in addressing a critical weak point in the Hadoop framework. The problem was the inability of active-active replication to work over wide-area networks (WAN). To solve this, WANdisco launched the Non-stop NameNode product, which eliminates the single point of failure (SPOF) in Hadoop systems.

The Non-stop NameNode enables active-active replication of the Hadoop NameNode, traditionally a critical component that manages the filesystem namespace and metadata but exists as a single active instance, making it a SPOF. If the NameNode fails, the entire Hadoop cluster can become unavailable.

Here’s how WANdisco’s Non-stop NameNode works:

  1. Active-Active Replication: Instead of having one active NameNode and a standby, WANdisco enables two or more NameNodes to run simultaneously in an active-active mode. Each NameNode can process client requests in real-time.
  2. Strong Consistency: WANdisco’s patented Distributed Coordination Engine (DConE) ensures that all active NameNodes remain perfectly synchronized with strong consistency guarantees. Changes made on one NameNode are immediately and atomically replicated to others, preventing split-brain scenarios and data inconsistencies.
  3. Zero Downtime Failover: Because multiple NameNodes are active and synchronized, if one NameNode goes down, clients automatically connect to another active NameNode without any interruption in service. This eliminates downtime caused by failover.
  4. Geographic Resilience: The technology can span multiple data centers or cloud regions, protecting the NameNode metadata service against site-level failures as well.

In June 2012, WANdisco launched on the Alternative Investment Market (AIM) of London's stock exchange, raising £15 million in the IPO. The company reported sales for 2012 of $6 million, up 56% from 2011, and following the announcement, WANdisco’s share price rose 3%.

In November 2012, WANdisco acquired AltoStor, a Silicon Valley big data storage start-up whose founders were two of Hadoop’s original authors. This acquisition combined AltoStor's Hadoop know-how with DConE, further strengthening WANdisco's position in the big data market.

WANdisco's core technology, the Distributed Co-ordination Engine (DConE), allows multiple instances of the same application to run on separate infrastructure, with data replicated in real time. By implementing code-sharing applications such as CVS and Subversion on DConE, businesses can make development more effective and resilient.

In a significant milestone, WANdisco has won its first Non-stop NameNode customer, an unnamed tier 1 telecommunications provider in the UK. This marks a significant step forward in the adoption of WANdisco's technology in the big data industry.

Dr Yeturu Aahlad, the brain behind WANdisco, left his job at Sun Microsystems in 2000 to work on a computer science problem. He solved the problem by applying the Paxos algorithm, a method devised by Leslie Lamport in the 1980s. This algorithm forms the backbone of WANdisco's Non-stop NameNode technology.

In 2009, WANdisco decided to base its new headquarters in Sheffield, citing lower costs and a significant talent pool in the North of England. The first commercial opportunity WANdisco pursued was distributed software development.

WANdisco made a £3 million loss for the year 2012, including £2.5 million in costs associated with the floatation. However, the company's innovative solutions and strategic acquisitions indicate a promising future in the big data landscape.

In February 2013, WANdisco announced Non-stop NameNode, a product that claims to make Hadoop systems impervious to NameNode outages. This announcement marks a significant step forward in WANdisco's mission to revolutionise the big data industry.

Technology plays a pivotal role in WANdisco's success in data-and-cloud computing, as demonstrated by the development and implementation of their Non-stop NameNode product. This revolutionary technology, a significant advancement in the Hadoop framework, ensures strong consistency, active-active replication, zero downtime failover, and geographic resilience, thereby mitigating potential weak points in big data systems.

Read also:

    Latest