Data Management (DMS) is a one-stop platform for managing the full lifecycle of your data. It covers database connections, governance, development, low-code data processing, real-time transmission, and disaster recovery — all from a single interface.
Connect and manage databases across environments
DMS connects to a wide range of database engines without requiring separate tools for each:
Relational databases: MySQL, SQL Server, PostgreSQL, PolarDB for MySQL, PolarDB for PostgreSQL (Compatible with Oracle), PolarDB-X, OceanBase, Oracle, and Dameng (DM)
NoSQL databases: Redis, MongoDB, Memcache, Cassandra, and Graph Database (GDB)
Online analytical processing (OLAP) databases: AnalyticDB for MySQL and AnalyticDB for PostgreSQL
File and log storage: Object Storage Service (OSS) and Log Service
For the full compatibility matrix, see Supported database types and features.
Connect on-premises and third-party cloud databases. DMS integrates the free Database Gateway service to connect databases hosted on other cloud platforms or in on-premises data centers. For details, see What is Database Gateway?
Migrate databases with minimal downtime. DMS provides Data Transmission Service (DTS) to support migrations throughout the entire database lifecycle. For details, see What is DTS?
Manage instances at any level of granularity. Query, register, modify, enable, disable, and remove instances. Grant or revoke permissions at the instance, database, table, column, and row level. The metadata access control feature ensures only authorized users can reach specific instances and databases. For details, see Databases supported by DMS.
Classify and search data assets. Tag instances, databases, and tables with categories so administrators, developers, and O&M engineers can locate and manage data quickly. The graph feature supports fast cross-asset lookups. For details, see Use the asset category feature.
Govern data with fine-grained access control and compliance audit
Enforce design standards before issues reach production. DMS ships with built-in Alibaba development specifications and supports more than 200 schema design and review standards. Customize the rule set to match your organization's requirements, and catch risks such as missing primary keys early in the development cycle.
Control access at the row level. Grant permissions — Query, Export, Change, and Logon — on instances, databases, tables, columns, and rows. DMS automatically manages account lifecycles, so access is revoked promptly when team members change roles. For details, see Overview.
Maintain a complete audit trail. DMS records every permission change, data modification, and database access event, giving your team the evidence needed for internal audits and regulatory reviews. For details, see Use the operation audit feature.
Meet major compliance frameworks out of the box. DMS supports the Cybersecurity Law of the People's Republic of China, the EU General Data Protection Regulation (GDPR), the Sarbanes-Oxley Act (SOX), the Payment Card Industry (PCI) Data Security Standard (DSS), and the Health Insurance Portability and Accountability Act (HIPAA). Sensitive data is automatically identified and classified to prevent unauthorized access. For details, see Overview.
Keep databases stable during changes
Change data without locking tables. DMS executes data changes in a lock-free manner, creates a backup before each change, and generates rollback SQL if a change fails. For details, see Overview and Data tracking.
Monitor data quality across the pipeline. Configure quality rules for data generation, integration, processing, and consumption. DMS monitors and verifies data automatically at each stage. For details, see Check the data quality.
Catch risky SQL before it runs. The SQL review feature analyzes uploaded SQL statements against security rules, flags statements without indexes and non-conforming patterns, and provides optimization suggestions — reducing exposure to SQL injection attacks. For details, see SQL review.
Monitor database instance performance. DMS incorporates capabilities from Database Autonomy Service (DAS) to track instance performance metrics and surface anomalies before they affect users. For details, see View the performance details of a database instance.
Roll back exceptional changes. Identify data changes made within a specific time window and generate rollback SQL statements to restore the database to a known-good state. For details, see Data tracking.
Develop databases efficiently across teams
Work across multiple database engines from one tool. DMS handles engine-specific syntax and behavior differences, so developers can switch between databases without switching to separate clients.
Define custom workflows per environment. Configure distinct development processes for development, test, staging, and production instances. Changes in non-production environments require no approval; production changes go through an approval gate. This keeps development fast while protecting production data. For details, see Manage iterations and Manage security rules.
Generate test data at scale. Clone databases and populate them with synthetic data — random values, region names, virtual IP addresses — in bulk, so test environments reflect realistic production conditions. For details, see Generate test data.
Save and reuse SQL templates. Store frequently used SQL statements in a shared library for quick retrieval across the team.
Process data with low-code tooling
Batch data processing
DMS provides a low-code environment for offline data integration tasks. Configure complex scheduling properties with minimal coding — no deep engineering expertise required.
Key capabilities include:
Build offline data warehouses and optimize warehouse performance
Analyze and compute offline data at scale
Build data middle-end for large-scale enterprise data integration
Integrate heterogeneous data sources, including a virtual data lake over your enterprise data assets
For details, see Overview.
Streaming data processing
DMS uses DTS to read and write data with distributed concurrency, delivering throughput faster than Flink and improving data accuracy for real-time workloads.
Use cases include:
Build real-time data warehouses
Run real-time join queries across multiple tables
Upload and process real-time data
Process real-time reports and separate business logic from computing
Troubleshoot business issues in real time
For details, see What is ETL?
Visualized data development
DMS provides a drag-and-drop orchestration canvas that supports hybrid task flows across multiple database types. Built-in data permission management controls who can schedule and modify complex pipelines. O&M features — task operation logs, data lineage tracking, and monitoring — are included.
For details, see Overview.
Transmit data in real time
DTS provides three data transmission methods — data migration, data synchronization, and data subscription — to cover the full range of integration scenarios. Key capabilities:
Distributed architecture: Removes single-instance bottlenecks with parallel processing
Throttling: Limits the load DTS places on source databases to protect production workloads
Flexible scheduling: Set recurrence intervals as short as 5 minutes
Simplified O&M: Performance monitoring, end-to-end diagnosis, and alerts are built in
For details, see Overview of data migration solutions. For synchronization and change tracking, see Overview of data synchronization solutions and Overview of change tracking scenarios.
Protect data with Data Disaster Recovery
Data Disaster Recovery is a backup and restore module built into DMS. It is built on Apsara Distributed File System and is designed for enterprises that need enterprise-grade recovery point objective (RPO) and recovery time objective (RTO) targets without the overhead of a self-managed backup system.
Cost efficiency
Pay-as-you-go billing avoids large upfront investments and ensures you pay only for what you use
Compressed and dedicated backup formats significantly reduce storage footprint compared to raw dumps
Tiered storage automatically moves backup data to lower-cost storage media as it ages, making long-term archiving affordable
Performance
Real-time incremental backup captures in-memory logs continuously, achieving an RPO within seconds. The latency varies based on network conditions.
Parallel backup with adaptive sharding uses multiple threads and runs without locking the source database
Point-in-time recovery provides a calendar and timeline UI so you can select any restore point with precision; full and incremental backups let you restore a database within seconds
For details on restoring individual tables, see Restore data by database or table.
Security
Encrypted transmission and storage uses SSL and AES-256 encryption; select the encryption method when configuring a backup schedule
Bring Your Own Key (BYOK) lets you encrypt backup data with keys managed in Key Management Service (KMS)
Geo-redundancy replicates backup data across zones for additional protection
Authentication supports whitelist configuration, hotlink protection, user management through Resource Access Management (RAM), and custom authentication
Automated alerts notify you of backup errors, restore errors, and restore completion
For alert configuration, see Manage alert rules. For geo-redundancy setup, see Overview.
Ease of use
Setup in 5 minutes: Purchase a backup schedule, configure it, and run the first backup task in under five minutes
Fine-grained backup scope: Back up entire instances, individual databases, individual tables, or any combination
Single-table restore reduces RTO by targeting only the affected table rather than the full instance
Lifecycle management: Define custom rules to automatically archive, clean up, replicate, or redistribute backup data
Data Disaster Recovery vs. self-managed backup
| Dimension | Data Disaster Recovery | Self-managed backup |
|---|---|---|
| Cost | Pay-as-you-go; compressed formats reduce storage costs; tiered storage for archiving; no staffing overhead for maintenance | Large upfront investment; storage bound by disk capacity; manual bandwidth scaling during peak hours; multi-tier media raises O&M costs |
| Security | SSL and AES-256 encryption; resource isolation between users; geo-redundancy; BYOK via KMS; validation at any time; RAM-based user management; custom authentication supported | Requires separate scrubbing devices and black-hole policies; needs a standalone security mechanism |
| Ease of use | 5-minute setup; fine-grained scope; lifecycle automation; web-based GUI | Complex scripting required; limited flexibility; basic capabilities only |
| Performance | RPO within seconds; point-in-time recovery to any point; single-table restore to reduce RTO; streaming backup with no disk flush; unlimited BGP backbone bandwidth | Performance bottlenecks from multiple tools |
| Reliability | Apsara Distributed File System with designed durability of at least 99.999999999%; real-time integrity verification | Mixed-tool environments increase risk; hardware failures such as disk bad sectors can cause data loss |
| Scalability | Backs up ApsaraDB databases, databases on ECS instances, self-managed data centers, and third-party cloud platforms including AWS and Tencent Cloud; restores to source or alternative environments | Tied to specific environments; generally not scalable |
For specification guidance, see Specification requirements.