India’s Ecommerce Powerhouse Snapdeal Sees Immediate Payoff with Aerospike

• 分析与建模 - 大数据分析
• 应用基础设施与中间件 - 数据库管理和存储
• 基础设施即服务 (IaaS) - 云存储服务

• 电子商务

• 商业运营
• 销售与市场营销

• 库存管理
• 预测性维护
• 实时定位系统 (RTLS)

• 软件设计与工程服务
• 系统集成

Snapdeal’s business and platform model is anchored by an innovative system enabling sellers to list their products, manage inventory and make pricing changes in real time while shoppers can review and rate sellers on issues such as shipping, delivery and returns. But challenges on how to maintain their platform’s real-time performance as the business scaled up 200 times arose. Sellers need to push their updates in real-time, and consumers demand a highly responsive online experience. With every page click, Snapdeal combines the updates from shoppers and sellers to display the most relevant products, as well as rankings for all the sellers that are offering the product by price, delivery time, and customer satisfaction. To support its inventory and pricing system, Snapdeal initially deployed MongoDB NoSQL database servers with data in DRAM as a cache in front of MySQL. The Snapdeal application used write-through techniques to update information first in MySQL and then in MongoDB, and it processed reads from MongoDB. However, as the business scaled and more sellers made price adjustments on more products, the MongoDB response times shot up from 5 milliseconds to over a second compromising the consumers’ shopping experience and leading to lost revenue opportunities. Worse, price changes were not always reflected in real-time.

Snapdeal is one of the leading e-commerce platforms in India, empowering sellers across the country to provide consumers with a fully responsive and intuitive online shopping experience. The platform merges logistics subsystems with cutting-edge online and mobile payment models, offering a wide range of products from thousands of national, international, and regional brands. Snapdeal.com has a network of more than 20,000 sellers, serving over 20 million consumers. The platform is powered by the Aerospike database, which enables real-time big data processing capabilities. This allows Snapdeal to provide consumers with up-to-the-moment updates about product availability, pricing, and seller ratings, making buying decisions informed and convenient.

Snapdeal wanted a technology solution that could affordably scale as the business expanded. It required a high-throughput, low-latency system that could scale from 2 million to 150 million records, from 10,000 to 30,000 reads per second, and from 30 to 500 writes per second. Snapdeal also sought a technology solution that could affordably scale as the business expanded and did not want to use an expensive clustered relational database management system (RDBMS). For Snapdeal, the Aerospike in-memory NoSQL database provided several advantages over others. Most of the systems reviewed by Snapdeal engineers failed to meet at least one of the criteria: The existing MongoDB solution lacked predictable response times under high write loads, sharding was complex, and hardware requirements for scaling were cost prohibitive. Clustered RDBMS databases and the Terracotta BigMemory Max caching technology were too expensive. Redis did not have a server-side distribution mechanism at that point of time. Amazon Memcache and Amazon DynamoDB did not deliver predictable low latency. Couchbase replicas could only be used as backup copies, not to distribute load. Moreover, it required twice the number of servers as Aerospike for the same throughput. Aerospike performed with predictable low latency with 95-99% of transactions completing within 10 milliseconds—essential for enabling a responsive customer experience. Second, Aerospike had the highest throughput. Third, Aerospike delivered the highest price/performance—offering the lowest cost solution in terms of both hardware requirements and ease of operations.

• In-memory Aerospike database maintains sub-millisecond latency on Amazon Elastic Compute Cloud (EC2) while managing 100 million-plus objects stored in DRAM to support real-time dynamic pricing.
• Predictable low latency with 95-99% of transactions completing within 10 milliseconds—essential for enabling a responsive customer experience.
• Aerospike’s highly efficient use of resources enables Snapdeal to cost effectively deploy in Amazon EC2.

• Snapdeal scaled from 2 million to 150 million records.
• Reads per second increased from 10,000 to 30,000.
• Writes per second increased from 30 to 500.