Community Blog Alibaba Cloud Summit - Making a Data Warehouse that Integrates Distributed, Elastic Computing, and Cloud Computing

Alibaba Cloud Summit - Making a Data Warehouse that Integrates Distributed, Elastic Computing, and Cloud Computing

Think about this…by 2020, there will be 40 ZB of data in the world. Can AnalyticDB transform into the ultimate form of a data warehouse?

By Yuechang

Join us at the Alibaba Cloud Summit 2020 to learn more about our Cloud-Native Intelligent Database solutions!

As stated in Moore's law, the cloud has become universal in the fourth technological revolution. As a result, the data surrounding and stored on the cloud has grown rapidly. By 2020, there will be 40 ZB of data in the world. In 2015, there was already 8.5 ZB of data produced, with China accounting for 22% of it. With the development and application of technologies, such as the cloud and with increasingly durable data, cloud computing, and big data analysis have become rocket-fuel for enterprises and business alike. These features are ensuring that businesses fulfill their resource potential. Industries have also used big data analysis platforms to expose serious problems in the realm of databases. These issues are a burden for real-time decision-making, system structure design, data processing, and storage.

What kind of big data analysis platforms do enterprises need?

It goes without saying that data is important, but it is hard to say what kind of data platform and service users need. Where are the most valuable services that enterprises need for data warehouse solutions?

1) High Performance

No matter what the industry is, clients will require high performance. Performance is the starting point for all data platforms. It is possible to optimize performance and lower costs for PB-level data. Traditional data warehouses need to be discarded for progress to be made. Only by using the help of the cloud is it possible to achieve this level of resource productivity and cost reduction.

2) Real-Time

In the era of self-BI, users do not need to generate reports through IT, and data requirements need to be more responsive to real-time issues. Modern day users need to analyze new data that was produced hours (or a few minutes) ago.

3) High Efficiency

The number of Internet users and mobile users is saturating, the demographic dividend is shifting, and the incremental market has entered the stock market. As a result, the stock market is more competitive. In this competitive environment, most enterprises are chasing a simple, fast, and efficient way to maximize the value of traffic streams at the minimum cost to bring them new business points.

Alibaba Cloud AnalyticDB – Next-Gen Cloud-Native Data Warehouse Service

To meet these higher requirements for enterprises, AnalyticDB, the new-generation cloud-native data warehouse service, will release and overturn traditional data warehouses

Alibaba Cloud AnalyticDB integrates the advantages of distributed computing, elastic computing, and cloud computing. This new kind of data warehouse has made great breakthroughs in performance, real-time, efficiency, and scale. AnalyticDB supports parallel access on a larger scale, provides faster read and write capabilities, and implements smarter hybrid query and load management. It improves resource utilization and reduces costs, allowing users to focus more on business development and data value.

1) Ultra-large scale

Based on the strong-consistency RAFT protocol, the in-sync replica (ISR) mechanism has a real-time write performance of up to tens of millions per second and supports a maximum free-storage space of 100 PB.

2) Robust Performance

Lightweight index construction methods, distributed hybrid computing engines, and optimizers provide faster and more complex-SQL read/write capabilities. The latest white paper released by the AnalyticDB team states that its performance is 100 times superior to MySQL. AnalyticDB ranks first in worldwide TPC-DS and TPC-H performance results.

AnalyticDB refreshed the TPC-DS 10 TB performance results list once again and ranks #1 in the world for performance and costs. Compared with the world's last leading database service, the Spark optimized version, the overall performance improved by 29%, and the unit cost is 1/3 of the price of Spark.

AnalyticDB refreshed the TPC-H 30 TB performance results for the first time and ranks #1 worldwide for performance and costs. Compared with the previous world record held by Microsoft SQL Server 2019, the comprehensive performance increased by 290%, and the unit cost is 1/4 of the price. It has become China's first product to win the list.

3) Flexibility

Based on the storage and computing-decoupled architecture, the storage space can be expanded to 100 PB in seconds, and the compute nodes can be quickly added from 3 nodes to 5000 nodes. The AnalyticDB design perfectly reflects challenges and redundancies from the history of data warehouse products. It is designed to solve many problems leftover from issues caused by traditional data warehouses: costly, low flexibility, and troublesome operations management.

4) Aggregate Computing and Analysis

With the rapid development of mobile Internet and intelligence, a large amount of unstructured data has been generated. Understanding how to quickly mine the value of massive unstructured data is particularly important. AnalyticDB supports aggregate analysis, online analysis, and ETL calculation of structured and unstructured data for a collaborative online. The integration process of the database and big data platform has been achieved. This has helped clients build a data warehouse simply and quickly. It allows clients to stay focused on business development and business value enhancement.

Traditional data warehouse providers, such as open-source big data platforms like Hadoop, Hive, and Spark have seemingly formed a direct competitive relationship with AnalyticDB. Why are traditional data warehouse and BI increasingly becoming alienated? Why do people want to keep new data warehouses a secret?

Traditional data warehouses have left many bad impressions: being expensive, having low flexibility, and having troublesome O&M problems. Fortunately, AnalyticDB was released, using the cloud computing shell to justify the rationality of the existence of modern data warehouses. Will AnalyticDB become the ultimate form of a data warehouse? Let's wait and see!

Learn more about AnalyticDB at https://www.alibabacloud.com/product/analyticdb-for-mysql

Join us at the Alibaba Cloud Summit 2020 to learn more about our Cloud-Native Intelligent Database solutions!

0 0 0
Share on


398 posts | 81 followers

You may also like



398 posts | 81 followers

Related Products