Data Management (DMS) is a one-stop data management platform that allows you to manage the lifecycle of data. This topic describes the benefits of DMS and the Data Disaster Recovery module that comes with DMS.
Management of a full range of data assets
A full range of data assets:
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 more information, see Supported database types and features.
Databases from multiple sources: DMS integrates the free Database Gateway service to provide stable and efficient connections. You can use Database Gateway to connect databases that are hosted on other cloud platforms or on-premises databases to DMS securely at a low cost. For more information, see What is Database Gateway?
Tools for smooth database migration: DMS provides Data Transmission Service (DTS) to support database migration throughout the entire lifecycle. For more information, see What is DTS?
Database management: You can perform operations on database instances, such as querying data in an instance, registering an instance, modifying an instance, and granting permissions on instances, databases, tables, data columns, and rows. You can also perform management operations such as revoking permissions, disabling an instance, enabling an instance, and removing an instance. The metadata access control feature allows you to grant permissions on instances and databases to users. This way, only authorized users can access these instances and databases. For more information, see Databases supported by DMS.
Data classification: You can classify and manage instances, databases, and tables. This helps administrators, developers, and O&M engineers manage data in tables. For more information, see Use the asset category feature.
Quick data query: You can search for instances, databases, tables, features provided by DMS, and tickets in DMS. The graph feature allows you to retrieve data quickly.
Reliable data governance
Secure and credible development: The latest built-in Alibaba development specifications help you avoid potential risks caused by improper database designs such as missing primary keys. DMS supports more than 200 schema design and review standards. You can specify standards based on your business requirements.
Comprehensive fine-grained access control: You can manage permissions on instances, databases, tables, columns, and rows, and grant permissions, such as Query, Export, Change, and Logon, as needed. For more information, see Overview.
Automatic management of DMS accounts: DMS automatically manages the accounts and permissions of DMS users. This helps prevent data leaks when DMS accounts are not revoked in time after personnel changes.
Compliance audit: DMS maintains records and logs of operations performed on databases, including permission management operations, database changes, and access records. This helps enterprises prepare for compliance audits, including audits for enterprises seeking to go public and regular internal audits. For more information, see Use the operation audit feature
Compliance with multiple laws and regulations: DMS conforms to 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 can be automatically identified and classified. This prevents abnormal access to sensitive data. For more information, see Overview.
Guaranteed database stability
Controllable data changes: You can change data without locking tables, back up data before a data change, or roll back the data change after a data change fails. This ensures the stability of your databases during data changes. For more information, see Overview and Data tracking.
Data quality monitoring: You can configure data quality rules for data generation, data integration, data processing, and data consumption to monitor and verify data quality. For more information, see Check the data quality.
SQL statement review and optimization: DMS reviews uploaded SQL statements and provides optimization suggestions integrated into security rules to prevent SQL statements without indexes or non-conforming SQL statements, reducing the risk of SQL injection attacks. For more information, see SQL review.
Diagnosis of database performance: DMS incorporates some features of Database Autonomy Service (DAS) to help you keep track of database instance performance and ensure stable, secure, and efficient database services. For more information, see View the performance details of a database instance.
Rollback of exceptional changes: You can find exceptional data changes that are performed in a period of time and generate SQL statements to roll back the changes. For more information, see Data tracking.
Efficient database development
Support for various engines: You can develop databases based on a variety of database engines in DMS, without having to worry about the differences between engines.
Custom development process: You can design and use different development processes for database instances that serve different business purposes based on organization requirements. This ensures schema consistency between environments, such as the development environment, test environment, staging environment, and production environment. For more information, see Manage iterations and Manage security rules.
Agile and efficient development: The development staff can develop compliant databases based on the design standards. No approval is required for development processes in a non-production environment. Only development processes in a production environment need to be approved. This ensures data security while enabling efficient and collaborative development.
Auto building of test environments: DMS allows you to clone databases and generate large amounts of data such as random values, region names, and virtual IP addresses at a time to prepare test data with ease. For more information, see Generate test data.
SQL statement reuse: You can add and manage frequently used SQL statements to reuse SQL templates.
Low-code data processing
DMS provides batch data processing. For more information, see Overview.
A variety of scenarios: The batch data processing feature can be applied to different scenarios. For example, You can develop offline data warehouses, optimize data warehouse performance, analyze and compute offline data, build data middle-end, and integrate large amounts of data, including heterogeneous data.
Low requirement for expertise: Personnel can complete various offline integration tasks by performing simple configurations. You can also configure complex scheduling properties of a task. Only minimal coding is required.
Virtual data lake based on data assets: DMS allows you to build a virtual data lake for the data assets of your enterprise. You can use the data assets whenever you need them. This reduces the cost of O&M and data lake formation for your enterprise.
DMS provides streaming data processing. For more information, see What is ETL?
A variety of scenarios: The streaming data processing feature can be applied to different scenarios. For example, you can develop real-time data warehouses, perform real-time join queries on multiple tables, upload and process real-time data, process real-time reports, separate business from computing, and troubleshoot business issues in real time.
High efficiency and low latency: You can use DTS to obtain and write data. Distributed concurrent read and write operations are supported. The speed of obtaining and writing data is faster than that by using Flink. DTS also improves data accuracy.
DMS provides visualized data development. For more information, see Overview.
Hybrid orchestration: You can create a task flow that has various database types. This meets the requirements for complex workflows.
Complex task management and scheduling: DMS integrates the data permission management feature to manage and schedule complex tasks. This improves the security of data development.
Robust O&M mechanism: DMS provides multi-dimensional O&M capabilities by supporting features such as task operation logs, data lineage management, and monitoring.
Real-time data transmission
Multiple data transmission methods: The real-time data transmission feature of DTS provides the following data transmission methods: data migration, data synchronization, and data subscription. This meets various business requirements.
High performance: A distributed integration architecture is supported. This resolves the bottlenecks of single-instance database services. In addition, DMS provides the throttling feature to protect data sources as needed.
Simple O&M: DMS provides features such as performance monitoring, end-to-end diagnosis, and alerts to simplify O&M.
Flexible scheduling: You can configure scheduling properties as needed. The minimum recurrence of data transmission can be set to 5 minutes.
For more information, see Overview of data migration solutions, Overview of data synchronization solutions, and Overview of change tracking scenarios.
Data Disaster Recovery
Cost-effectiveness
Data Disaster Recovery uses Apsara Distributed File System as built-in storage. Backup data is converted to a dedicated format, compressed, and then saved to the built-in storage. This reduces storage costs.
Data Disaster Recovery supports the pay-as-you-go billing method to avoid investing a large amount of assets at a time.
In addition, Data Disaster Recovery supports tiered storage and automatically stores backup data in different storage media. You can archive backup data to reduce costs.
High performance
Real-time backup: Data Disaster Recovery captures in-memory logs in real time and achieves the recovery point objective (RPO) within seconds.
Data Disaster Recovery reads and parses database logs in real time by using the real-time data streaming technology of Alibaba Cloud. Then, Data Disaster Recovery backs up and stores data in the cloud to perform incremental backup. Data Disaster Recovery keeps the latency within seconds during incremental backup. The latency varies based on network conditions.
Parallel backup: Data Disaster Recovery can back up data in an unlocked manner, use multiple threads to back up data in parallel, and adaptively shard data during data pulling.
Restoration to a point in time: Data Disaster Recovery provides a calendar and a timeline so that you can select a point in time to which data is restored.
Data Disaster Recovery enables you to restore a database within seconds by using full backups and incremental backups. This ensures the security and integrity of your data. For more information, see Restore data by database or table.
Multiple specifications: Data Disaster Recovery provides scalable capabilities to seamlessly support the performance requirements of enterprises at different stages. For more information, see Specification requirements.
High security and reliability
Encrypted transmission and storage:
Data Disaster Recovery uses SSL and AES-256 encryption to secure backup data during transmission and storage. When you configure a backup schedule, you can select an encryption method based on your business requirements.
Data Disaster Recovery supports Bring Your Own Key (BYOK) so that you can encrypt backup data by using data keys generated by Key Management Service (KMS).
Automated alerting: Data Disaster Recovery sends notifications about key events such as backup errors, restoration errors, and restoration success. For more information, see Manage alert rules.
Geo-redundancy: This feature enhances the level of data protection. For more information, see Overview.
Flexibility and ease of use
Fine-grained backup: Data Disaster Recovery allows you to back up data of varying granularities based on your requirements, including entire instances, individual databases and tables, and multiple databases and tables.
Single-table restoration: Data Disaster Recovery allows you to select a single table as the object to restore. This reduces the recovery time objective (RTO).
Lifecycle management: Data Disaster Recovery supports custom rules for the lifecycle management of backup data. You can customize rules to automatically dump, clean up, duplicate, and distribute backup data.
Comparison between Data Disaster Recovery and self-managed backup systems
Item
Data Disaster Recovery
Self-managed backup system
Cost
The pay-as-you-go billing method ensures 100% resource utilization and avoids a large amount of upfront asset investment.
Cold data is separated from hot data for tiered storage. This is suitable for the long-term archiving of backup data.
The compressed and compact backup formats can significantly reduce storage costs.
No investments in hiring maintenance personnel or hosting databases are required.
A large amount of upfront asset investment is required.
Storage space is limited by hard disk capacity. The storage space must be manually increased.
Single-line or double-line access is slow, and bandwidth is limited. The bandwidth must be manually increased during peak hours.
The introduction of multi-level storage media leads to a sharp increase in O&M costs.
Security
Data Disaster Recovery uses SSL and AES-256 encryption to secure backup data during transmission and storage.
Resources are isolated between different users, and geo-disaster recovery is supported.
Data Disaster Recovery provides a variety of authentication and authorization methods, such as whitelist configuration, hotlink protection, and user management based on Resource Access Management (RAM).
Data Disaster Recovery allows you to verify the validity of backups at any time, and notifies you of the task status.
Custom authentication is supported.
Additional scrubbing devices and black hole policy-related services are required.
A separate security mechanism is required.
Ease of use
The process of purchasing and configuring a backup schedule and running a backup task takes only 5 minutes.
Fine-grained backup is supported. You can back up data of different granularities based on your business requirements, including an entire instance, multiple databases, a single database, multiple databases, or a single table.
Data Disaster Recovery supports global rules for the lifecycle management of backup data. You can customize rules to automatically dump, clean up, duplicate, and distribute backup data.
Data Disaster Recovery provides a web-based GUI for you to perform backup and restore operations with ease.
The backup process requires complex scripts and tools, which are difficult to learn.
A self-managed backup system is not flexible and provides only basic capabilities in most cases.
Performance
Data Disaster Recovery captures in-memory logs in real time and achieves a recovery point objective (RPO) within seconds. Data Disaster Recovery allows you to restore backup data to any point in time.
Data Disaster Recovery allows you to select a single table as the object to restore. This greatly reduces the recovery time objective (RTO).
Data Disaster Recovery supports streaming backup. Data is not flushed to disks. The entire backup window is unlocked. The backup speed can be adjusted based on the concurrency configuration.
Data Disaster Recovery uses a multi-line Border Gateway Protocol (BGP) backbone network that has no bandwidth limits. This allows a large number of users to perform backup and restore operations simultaneously.
The shortcomings of multiple tools used for backup are performance bottlenecks.
Reliability
Data Disaster Recovery uses Apsara Distributed File System to provide a distributed storage service with high reliability.
Data Disaster Recovery uses a redundant storage design to deliver a designed durability of at least 99.999999999%.
During the backup process, data integrity is verified in real time.
Tested by a large number of users, Data Disaster Recovery can efficiently detect and fix vulnerabilities.
The mixed use of multiple tools causes high risks.
A self-managed backup system is prone to errors due to low hardware reliability. If a disk has a bad sector, data may be lost.
Scalability
Data Disaster Recovery allows you to back up ApsaraDB databases and databases that are deployed on ECS instances, in self-managed data centers, or on third-party cloud platforms such as Amazon Web Services (AWS) and Tencent Cloud.
In addition to restoring data to the source database, Data Disaster Recovery also allows you to restore backup data to other environments. For example, you can restore an on-premises database to an Alibaba Cloud database by using Data Disaster Recovery.
Self-managed backup systems support only specific environments and are generally not scalable.