What is AnalyticDB for MySQL? - AnalyticDB - Alibaba Cloud Documentation Center

AnalyticDB for MySQL is a fully managed data warehouse service built on a data lakehouse architecture. It processes petabytes of structured, semi-structured, and unstructured data in real time, updates data within milliseconds, and responds to queries within sub-seconds.

The service is highly compatible with MySQL, so it integrates with your existing clients, drivers, business intelligence (BI) tools, and scheduling workflows without changes.

What you can do with AnalyticDB for MySQL

AnalyticDB for MySQL supports four major workloads. Each workload is self-contained — use one or combine them in a single cluster.

Multi-source data integration

Ingest structured, semi-structured, and unstructured data from over a dozen source types, including relational databases, NoSQL databases, cloud storage, log systems, and message queues. All data lands in a unified lakehouse that feeds your analytics and batch pipelines. For the full source list, see Data import.

Real-time analytics

Synchronize data from operational databases and streaming sources — including PolarDB, ApsaraDB RDS, ApsaraMQ for Kafka, and Simple Log Service — in real time. Data becomes queryable milliseconds after it is written, with strong consistency guaranteed.

Batch ETL processing

Extract data from sources, transform and cleanse it, and load it into AnalyticDB for MySQL on a schedule. Scheduling tools including Data Management (DMS), DataWorks, Airflow, DolphinScheduler, and Azkaban integrate directly with the cluster.

Spark-based data analysis

Run complex analytics and machine learning directly on lakehouse data using the built-in Spark compute engine. Use Spark SQL for structured queries, Spark JAR packages for custom batch jobs, or PySpark for preprocessing and ML pipelines.

Key capabilities

Supported data sources

AnalyticDB for MySQL ingests data from relational databases, NoSQL stores, big data platforms, object storage, log systems, message queues, and on-premises files.

Type	Data source
Relational database	ApsaraDB RDS for MySQL, ApsaraDB RDS for SQL Server, PolarDB for MySQL, PolarDB-X, self-managed MySQL database, self-managed Oracle database
Non-relational database	ApsaraDB for MongoDB, Lindorm, self-managed HBase database
Big data	MaxCompute, Flink, Hive
Storage	Object Storage Service (OSS), AWS S3, Azure Blob Storage, Google Cloud Storage, Tablestore, HDFS
Log	Simple Log Service, Logstash
Message queue	Kafka
On-premises file	Text file

For setup instructions, see Data import.

Ecosystem compatibility

AnalyticDB for MySQL is compatible with MySQL, so it works with most clients, drivers, BI tools, and scheduling tools that support MySQL — with no additional configuration.

Category	Tools
Client	DBeaver, DBVisualizer, Navicat, SQL WorkBench/J, MySQL command-line client
Driver	Java: Java Database Connectivity (JDBC) and Druid connection pool · Python: MySQLdb · PHP: MySQLi and PDO · C#: MySql.Data · Go: go-sql-driver/mysql · Node.js: MySQL
BI tool	FineBI, FineReport, Power BI, Tableau, Quick BI, DataV, , , QlikView, Smartbi, Superset, Metabase, Redash
Scheduling tool	DMS: Job scheduling · DataWorks: XIHE SQL jobs and Spark jobs · Airflow: XIHE SQL jobs and Spark jobs · DolphinScheduler: XIHE SQL jobs and Spark jobs · Azkaban: XIHE SQL jobs and Spark jobs

High performance

AnalyticDB for MySQL handles complex queries in seconds or milliseconds — delivering up to 10x the performance of relational databases for analytical workloads. Data becomes queryable milliseconds after being written, with strong consistency guaranteed.

Elastic scaling

AnalyticDB for MySQL uses a storage-compute decoupled architecture, so compute and storage scale independently.

Compute: Scale manually within predefined specifications, or configure automatic elastic scaling to handle business fluctuations. See Scale a cluster.
Storage: Scales automatically based on data volume, billed on a pay-as-you-go basis. Store historical data in OSS to reduce costs.

Security and compliance

AnalyticDB for MySQL covers permissions, network access, encryption, auditing, and backup — the full security surface for enterprise workloads.

Category	Feature	Description
Permission management	Resource Access Management (RAM)	Grant RAM users scoped access to create, manage, and delete clusters.
Permission management	Database permission management	Assign permissions at the global, database, table, or column level.
Data security	IP address whitelists	Block all inbound connections by default; allowlist only the IPs that need access.
Data security	SSL encryption	Encrypt connections at the transport layer using certificates issued by certificate authorities (CAs).
Data security	Disk encryption	Encrypt data at the block storage level so that it cannot be decrypted even if data leaks occur.
Audit	SQL audit	Log all DML and DDL operations in real time to detect anomalies and diagnose performance issues.
Backup and restoration	Periodic backup	Back up cluster data automatically on a configurable schedule.
Backup and restoration	Data restoration	Restore data from backup sets to a new cluster.
Monitoring and alerting	Cluster monitoring	View CPU utilization, compute memory usage, disk usage, and response time in the console or via API.
Monitoring and alerting	Spark monitoring	View Spark job metrics in the CloudMonitor console or via API.
Monitoring and alerting	Alerting	Configure alert rules so that the right contacts are notified when a threshold is breached.

For the full feature list, see Features and capabilities.

Pricing

Fees consist of reserved resource fees, elastic resource fees, data storage fees, cache storage fees, and backup storage fees.

For billable items, see Billable items of Enterprise Edition and Basic Edition.
For prices, see Pricing for Enterprise Edition and Basic Edition.

Editions

AnalyticDB for MySQL is available in two editions with the same feature set but different storage architectures.

Edition	Architecture	Nodes	Best for
Enterprise Edition	Multi-replica	Reserved nodes in multiples of 3	Production environments
Basic Edition	Single-replica	One reserved node	Learning and testing

To get started, create an Enterprise Edition or Basic Edition cluster.

Get started

New to AnalyticDB for MySQL? Start with creating a cluster, then follow the role-based guides below.

Database administrators

Manage accounts and permissions — Set up database users and control what each account can access.
Configure IP address whitelists — Restrict which clients can connect to the cluster.
Monitor cluster performance and workloads — Track CPU, memory, disk, and query response time.
Configure SQL audit — Log DML and DDL operations for compliance and anomaly detection.
Configure backup cycle and frequency — Schedule automatic backups to prevent data loss.

Data development engineers

Create databases and tables — Define the schema for your data warehouse.
Import data from warehouses or lakes — Migrate or synchronize data into the cluster. See also data lake ingestion.
Connect via application code — Use the JDBC, Python, Go, or other drivers to run queries from your services.
Design table schemas for query performance — Optimize partitioning and indexing for your workload.
Run Spark SQL or Spark applications (Spark apps) — Cleanse, transform, and compute data at scale.

Data analysts

Connect with a client — Use DBeaver, Navicat, or another MySQL-compatible client.
Build dashboards with BI tools — Connect Tableau, Power BI, Quick BI, or another supported BI tool.
Query data with SELECT statements — Run analytical queries on petabyte-scale datasets.
Use built-in functions — Apply aggregation, string, math, and other functions in your queries.
Run full-text search — Perform fuzzy match and similarity search across text data.
Optimize query performance — Diagnose slow queries and apply tuning recommendations.

Algorithm engineers

Preprocess data with PySpark — Cleanse, transform, join, and union large datasets for ML pipelines.
Run machine learning prediction with SQL — Invoke ML models and surface predictions directly in SQL queries.