Data centre's new chosen operating system: Hadoop, as stated by WANdisco
WANdisco's Non-stop NameNode product aims to improve the reliability and availability of Hadoop systems by addressing the critical issue of single points of failure (SPOF) inherent in the Hadoop NameNode architecture. This innovative solution offers several key applications and benefits that make it an attractive choice for enterprises.
Applications
- High Availability for Hadoop NameNode: The product provides active-active replication of the Hadoop NameNode, ensuring continuous operation without downtime, even during failover or maintenance.
- Disaster Recovery and Business Continuity: WANdisco's solution ensures that metadata changes in the Hadoop cluster are continuously replicated across geographically dispersed data centers. This protects data and metadata from data center outages or disasters.
- Multi-site Hadoop Deployments: Organizations running multiple Hadoop clusters across different sites can use WANdisco Non-stop NameNode to keep NameNodes perfectly synchronized, simplifying management and data consistency.
- Cloud and Hybrid Deployments: It supports deployments across on-premises and cloud environments, enabling hybrid cloud architectures with continuous metadata synchronization and failover capabilities.
Benefits
- Elimination of Single Point of Failure: Traditional Hadoop NameNode architecture can cause downtime if the active NameNode fails. WANdisco’s Non-stop NameNode creates a highly available active-active setup that prevents such failures from causing interruptions.
- Zero Downtime Failover: The solution enables seamless failover between NameNodes without any downtime or loss of metadata changes, ensuring uninterrupted Hadoop cluster operation.
- Real-time Metadata Synchronization: Unlike standby NameNodes that replicate logs asynchronously, WANdisco’s technology uses patented active-active replication, ensuring real-time consistent metadata replication.
- Improved Performance and Scalability: By allowing multiple active NameNodes, the system can serve metadata requests more efficiently, potentially improving performance in large or distributed Hadoop deployments.
- Simplified Disaster Recovery: Continuous replication across multiple sites ensures that the Hadoop environment can quickly recover from disasters without significant data loss or manual intervention.
- Operational Flexibility: IT teams can perform maintenance, upgrades, or hardware replacements without taking the Hadoop NameNode offline, reducing scheduled maintenance windows and operational risk.
Success Stories and Milestones
In a significant milestone, WANdisco has won its first Non-stop NameNode customer, a tier 1 telecommunications provider in the UK. The company's core technology, the Distributed Co-ordination Engine (DConE), has been instrumental in solving the problem of active-active replication not working over wide-area networks (WAN), a problem first identified by Dr Yeturu Aahlad in 2000.
WANdisco's technology has attracted customers such as HP, Intel, and Lockheed Martin. The company, founded in 2005 by Aahlad and UK IT entrepreneur David Richards, went public in 2012, raising £15 million in the IPO. Following the announcement of its first Non-stop NameNode customer and 2012 sales, 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 while at Yahoo! This acquisition has allowed WANdisco to combine AltoStor's Hadoop know-how with DConE, addressing a critical weak point in the big data framework.
In summary, WANdisco’s Non-stop NameNode product offers an enterprise-grade solution for high availability, disaster recovery, and zero-downtime operations. It ensures Hadoop clusters remain operational, consistent, and performant, protecting enterprises against costly interruptions caused by NameNode failures.
Technology and data-and-cloud-computing solutions like WANdisco's Non-stop NameNode product leverage innovative architecture to enhance the reliability and performance of Hadoop systems. This technology supports applications such as active-active replication of Hadoop NameNode for continuous operation, disaster recovery, multi-site Hadoop deployments, and cloud and hybrid deployments.
