×
Community Blog The Alibaba Cloud Database Technologies behind Double 11 in 2022

The Alibaba Cloud Database Technologies behind Double 11 in 2022

This article discusses how (and which) Alibaba Cloud technologies empowered Double 11 in 2022.

>> The Alibaba Cloud 2022 Double 11 Mega Sale is live! Enjoy a 75% discount with Data Transfer of Live, DCDN from $1 per year!

Alibaba Cloud cloud-native databases support the Double 11 Shopping Festival in 2022 with core technologies (such as serverless, HTAP, data lakehouse, and multi-model integration). Alibaba Cloud built a flexible, efficient, and green Double 11 Shopping Festival, helped e-commerce platforms and merchants reduce costs and improve efficiency, and brought consumers a better shopping experience.

PolarDB Managed Tens of Billions of Orders

The JUSHUITAN SaaS Enterprise Resource Planning (ERP) System Managed Nearly One-Fifth of Packages in China during Double 11 in 2022.

1

Shanghai JUSHUITAN Network Technology CO., LTD. is a leading SaaS ERP company in the e-commerce industry in China.

JUSHUITAN has served more than 800,000 merchants. Currently, the JUSHUITAN SaaS ERP system manages one-fifth of packages delivered by e-commerce merchants in China.

PolarDB supports hybrid transaction/analytical processing (HTAP) implemented based on the In-Memory Column Index (IMCI) feature. HTAP helps JUSHUITAN improve its core business system. PolarDB serves more than 20,000 leading merchants and assists in data-driven business operations.

Before Double 11, JUSHUITAN added column store nodes to the existing PolarDB cluster. This improves the data analysis capabilities of the JUSHUITAN SaaS ERP system, enables integrated online transaction processing (OLTP) and online analytical processing (OLAP), and facilitates product launch and commercialization. The elastic scaling capability of PolarDB resolves the challenges of high concurrent queries per second (QPS) caused by the growing number of merchants that JUSHUITAN serves and helps JUSHUITAN handle burst traffic during Double 11.

PolarDB-X Improved the Shopping Experience

PolarDB-X Processed Tens of Millions of Concurrent Requests per Second (RPS) and Handled the Business during Peak Hours Easily during Double 11 in 2022.

2

PolarDB-X ensures an excellent shopping experience in the face of a large number of orders and ultra-high concurrent requests.

In core e-commerce transaction scenarios, databases need to process tens of millions of concurrent RPS and have high availability across regions during Double 11.

PolarDB-X ensures strong data consistency based on enhanced Paxos that requires data storage in at least three replicas and provides active geo-redundancy in a cell-based architecture. Based on distributed linear scalability, PolarDB-X is used to build a distributed cluster that contains thousands of servers. This cluster can process tens of millions of concurrent transactions per second (TPS), which prevents service degradation for the peak traffic at 00:00 on Double 11.

AnalyticDB Made Analysis and Decision-Making More Efficient

AnalyticDB Helped Merchants Make Advertising Decisions within Seconds during Double 11 in 2022.

3

AnalyticDB provides features (such as data lakehouse and cache acceleration) to help merchants make decisions about real-time precision marketing.

Merchants can use databanks to make marketing decisions. AnalyticDB supports databanks to integrate petabytes of data and store data in a data lake. This breaks data silos. Databanks manage the integrated data and support multiple analysis systems, which reduces the flow of large amounts of data. The technologies of AnalyticDB (such as bitmap index and cache acceleration) are suitable for core scenarios (such as real-time precision marketing and advertising). These technologies allow merchants to make decisions within seconds. The time consumed for data analysis is reduced from days to hours, and the overall cost is reduced by 40% each month.

Lindorm Made Logistics Faster

Cainiao Reduced the Cost of Promotional Resources by More Than 20% during Double 11 in 2022.

4

Lindorm helps online services (such as Cainiao) and order fulfillment platforms improve efficiency and reduce costs. The core technologies of Lindorm (such as storage pooling and dictionary encoding) are conducive to reducing carbon emissions and making Double 11 greener.

Lindorm provides core features (such as storage pooling, hot and cold data separation, and dictionary encoding) in the context of advocating emission reduction and energy conservation. These features help Cainiao reduce the cost of promotional resources for hundreds of terabytes of data by more than 20% each year.

Lindorm is one of the key databases for big data scenarios and has been supporting Double 11 for many years. Lindorm ensures the stable operation of core scenarios (such as Taobao Mobile, Intelligent Chatbot, data dashboards, and pay-per-click advertising). Lindorm provides multiple retrieval capabilities and uses standard SQL statements to improve business development efficiency. Lindorm distributes data across three zones within the same region to achieve disaster recovery with strong consistency across data centers within seconds. This ensures data security and stability during Double 11.

ApsaraDB RDS Helped ISVs Manage Resources

Independent Software Vendors (ISVs) Could Scale Resources up or down within Five Seconds during Double 11 in 2022.

5

ApsaraDB RDS provides Serverless capabilities to help merchants easily handle traffic fluctuations during Double 11.

A large number of Taobao and Tmall ISVs prefer to use ApsaraDB RDS databases as their business base. Serverless ApsaraDB RDS for MySQL databases ensure database stability and performance. In addition, Serverless ApsaraDB RDS for MySQL databases use innovative technologies (such as high-performance elasticity and second-level billing) to help merchants automatically respond to traffic loads and improve the utilization of cloud resources, thereby reducing costs. Compared with the preset storage capacity for peak hours, the use of Serverless ApsaraDB RDS for MySQL databases can reduce costs by more than 70%.

ApsaraDB RDS provides high-performance, stable, secure, and auto-scaling online database services and a complete set of solutions for disaster recovery, backup and restore, monitoring, and migration. This helps troubleshoot issues that may occur while using ApsaraDB RDS databases.

Tair Made Shopping Carts “Bigger”

Consumers Can Add up to 300 Items to Their Shopping Carts during Double 11 in 2022.

6

Tair is a cloud-native in-memory database service. Tair helped the e-commerce industry build the first large-capacity shopping cart. This prevents issues caused by shopping cart abandonment.

Shopping carts used by hundreds of millions of consumers bring a huge network overhead during Double 11, requiring a cost-effective database that supports a large number of connections and high throughput.

Tair has high-performance read and write speeds. Based on the connection aggregation and asynchronous write technologies, a single Tair cluster can support the traffic of 100 GB/s, which supports a large number of connections. The latest real-time processing technologies (such as executor and operator) optimize the shopping cart scenario, so the system latency is kept below 0.8 ms. This allows consumers to see the prices of commodities on the shopping cart page.

DMS Manages and Protects Data

Data Management Service (DMS) Identified and Protected More Than 170 Types of Business Data during Double 11 in 2022.

7

DMS uses machine learning technologies to identify and protect more than 170 types of business data, including order information, customer information, and product information.

DMS scans e-commerce links and provides all-around data protection during Double 11.

DMS protects business data by implementing access control on resources, intercepting high-risk operations, performing data masking, adding data watermarks, and performing tamper-proofing audits. This ensures data can be accessed as usual (even during the peak hours of Double 11). DMS also helps customers in industries (such as the Internet, finance and securities, automobile, and telecommunication) identify and protect business data for over 100,000 users.

0 0 0
Share on

ApsaraDB

258 posts | 20 followers

You may also like

Comments