Community Blog Introducing Data Lake Analytics: Serverless, Cloud Native, and Zero Upfront Cost

Introducing Data Lake Analytics: Serverless, Cloud Native, and Zero Upfront Cost

Alibaba Cloud Data Lake Analytics (DLA) is an interactive serverless analytics service that allows you to analyze your data at extremely low costs.

Alibaba Cloud Data Lake Analytics (DLA) is a serverless, high performance interactive query service that requires zero infrastructure setup or upfront cost. Customers can use DLA to query data stored in a number of Alibaba Cloud services, such as objects in OSS, key-value pairs in Table Store, and relational data in RDS. Customers can explore a single dataset at a time or analyze data across multiple sources simultaneously, and DLA delivers fast, interactive responses with a Massive Parallel Processing (MPP) architecture. DLA is fully compatible with SQL syntax which customers are already familiar with and therefore have an extremely low learning curve. In addition, DLA works with popular BI tools natively so that customers can easily visualize their data, derive insights, and accelerate decision-making.

Fig.1: Data Lake Analytics in Alibaba Cloud

Target Users

DLA is designed for those who need cloud-native analytics and are seeking a cost-effective solution that can turn raw data into real-time insights. Compared with conventional analytics platforms which require upfront hardware provisioning and setup, DLA allows customers to run queries on demand with no upfront costs and quickly explore ideas on new data while only paying for the resources consumed. Designed for developers, business analysts, and data scientist where a self-service operational model is highly preferred, DLA provides built-in ETL (Extract, Transform and Load) functionality that reduces the heavy lifting of pre-processing data of different formats and origins so customers can focus on analysis and insights.


Serverless: For enterprise users, the service is made available on a Pay-As-You-Go basis with ZERO maintenance effort. Key benefits include but not limited to instant startup, transparent upgrade, and elastic QoS.

Database-like User Experience: DLA offers standard SQL interface, with outstanding SQL compatibility and comprehensive built-in functions. Based on JDBC/ODBC connectivity, users are granted with fast and convenient service access, as well as low migration cost. Meanwhile, integration with BI product enables DLA to turn big data into consumable insights and visualizations. With database experience, DLA helps customers accelerate their cloud migration process.

Deep Ecosystem Integration: DLA enables complex analytics for data coming from different sources with various formats. Not only user can leverage DLA to analyze data stored on Alibaba cloud OSS and Table Store respectively, but also, they can join the results and thus generate new insights.

Optimized Performance: DLA fully leverages Massive Parallel Processing (MPP) architecture, vectorized execution optimization, multi-tenancy resource allocation and priority scheduling to achieve optimized performance.

Key Usage Scenarios

For raw data (logs, CSV, JSON, Avro, etc.) persisted on a given storage, OSS for example, you can query specified objects/files/folders without having to go thru a complex ETL process, and get the results back instantly

Fig.2: Query OSS Data

When dealing with multiple data sources, OSS and Table Store for example, DLA enables JOIN operation across heterogeneous data sources, turning your big data into consolidated insights.

Fig.3: Joint Analytics Across OSS and Table Store


Data Lake Analytics is now available in the Singapore, China (Hangzhou, Shanghai, Beijing), and the UK regions, and would be coming soon to US (West) and Australia regions. To get started on Data Lake Analytics, watch the DLA webinar here or visit the product page here.

Check out the latest offers for Data Lake Analytics at www.alibabacloud.com/campaign/data-lake-analytics-2018

1 0 0
Share on

Alibaba Clouder

2,606 posts | 737 followers

You may also like


Raja_KT November 20, 2018 at 3:49 pm

I hope we can join together data from HDFS, OSS, Tablestore, RDBMS and CPU spike will not be there for UDFs in case it is there.Also, I see "You can only analyze files on OSS/Table Store which is located in the same region where the select DLA service is deployed. Cross-region operation is not yet supported."