ADSQL for MySQL
A high performing shard-enabled distributed database highly compatible with MySQL
Overview
Benefits
Features
Overview
ADSQL for MySQL (ADSQL) presents a comprehensive distributed database solution renowned for its robust consistency, high availability, and performance. It boasts a global deployment architecture, enabling distributed horizontal scaling while upholding enterprise-grade security measures. Additionally, ADSQL encompasses value-added services such as intelligent database administration, automated operations, and robust monitoring capabilities. Currently, it caters to over 500 governmental and organizational clients spanning various sectors, including banking, insurance, securities, internet finance, billing, third-party payment, IoT, Internet+, and government affairs, delivering database services in both public and private cloud environments. ADSQL has garnered numerous international and national accreditations, along with commendation from satisfied customers and industry stakeholders, underscoring its reputation for excellence in product quality and service provision.
Benefits
Automated Sharding
By setting a shard key when creating a table, you can automatically split a large table horizontally (i.e., sharding) in the database. The ADSQL system will then distribute the data evenly across different physical shards using a hash scheme, and it will automatically aggregate queries to return results.
High Compatibility with MySQL Syntax
ADSQL is compatible with commonly used MySQL syntax elements, including language structures, character sets, time zones, data types, common functions, preprocessing protocols, sorting, joins (JOIN), indices, partitions, transactions, control commands, and typical DDL, DML, and DCL commands, as well as database access APIs.
Leading Distributed Architecture
Given the challenges of balancing advanced functionality and high performance in distributed databases, ADSQL offers three table creation solutions to address this issue. These solutions feature distributed transaction characteristics, globally unique numerical sequences, and JSON format support, effectively bridging the gaps in distributed architecture and enabling more flexible development options.
Elastic Scalability Without Downtime
Currently, a single shard can support up to 6 TB of storage. If additional performance or capacity is needed, you can effortlessly scale up with just a few clicks in the console. There is no need to worry about data migration, balancing, or route switching within the distributed system, as the IP remains unchanged and the service will only be interrupted for a few seconds during the automatic switch.
Strong Synchronous Replication
By default, ADSQL uses a master-slave architecture that ensures over 99.95% availability. It supports strong synchronous replication for robust data consistency, meaning an application transaction is acknowledged only after the data written by the business system is synchronized to the slave. This guarantees complete master-slave data consistency and eliminates the risk of data loss or corruption.
Ultra-high Performance
Built on a deeply customized and optimized MySQL kernel, ADSQL significantly outperforms MySQL. It supports read-write separation through three different schemes, ensuring development flexibility and enhancing read scalability. Additionally, it features optimized thread pool scheduling algorithms that deliver superior performance under heavy loads.
Convenient Managed Deployment
A production-ready ADSQL instance can be launched and connected within minutes with just a few clicks in the ADSQL console. The console includes commonly used database operations, allowing for refined database management. Additionally, it displays essential system monitoring and performance analysis data, helping you quickly identify and resolve database issues.
JSON Support
ADSQL supports native JSON types by storing JSON values in an internal binary format rather than as strings. This enables rapid reading of document elements and ensures automatic validation of JSON texts upon insertion or update in JSON columns. Essentially, ADSQL functions as a NoSQL database that is compatible with the MySQL protocol.
RocksDB Support
RocksDB, a storage engine based on the LSM tree, offers a data compression rate that is over 60% higher than that of the InnoDB engine. Typically, it requires only about 1 GB of storage space for 10 million rows of data after compression, significantly lowering database usage costs. Additionally, it excels in high write performance, making it ideal for space-sensitive business scenarios with more writes than reads, such as IoT and logging.
Features
High Compatibility with TDSQL
Enhanced Performance
High Availability
Enterprise-grade Security
Enterprise-grade Monitoring
Scalability
Ease of Use
Compatible with frequently used MySQL syntax elements
These include MySQL’s language structures, character sets, time zones, data types, common functions, preprocessing protocols, sorts, joins (JOIN), stored procedures, indices, partitions, transactions, control commands as well as common DDL, DML, and DCL commands and database access APIs.
(Distributed) Transaction Support
ADSQL facilitates distributed transactions via a two-phase commit mechanism, particularly suited for applications like money transfer, billing, and payments. Its cross-node transactional performance reaches approximately 70% of a single node, surpassing XA, an open-source distributed transaction protocol, by approximately 56%. Presently, this functionality is solely available on MySQL 5.7.
(Distributed) JOIN Support
ADSQL supports JOIN (joint query) operations across physical nodes, namely, distributed joins. If a join-related table has a shardkey equality condition, thanks to the consistency principle of sharding, this part of data will be automatically stored to the same physical node. In this case, it is equivalent to performing a standalone join which has the best performance. If cross-physical node data is involved, the proxy will first pull data from other nodes and cache it, in which case performance will be compromised due to network data transfer.
JSON Support
ADSQL supports the storage of JSON formatted data, enhancing the efficiency of JSON processing and facilitating early error detection. Opting for ADSQL allows you to leverage JSON types without sacrificing traditional database capabilities like data consistency, transactions, and joins. ADSQL’s JSON support, rooted in MySQL, offers a slightly different usage experience compared to MongoDB. For those interested, a comparison of JSON capabilities between ADSQL and MongoDB is available for reference.
Three Table Creation Solutions
Sharded Table: A horizontally partitioned table typically employed for managing large and rapidly expanding datasets, typically exceeding 20 million rows or 50 GB in size.
Broadcast Table: All operations are propagated to all logical shards, meaning each shard contains the complete dataset of the table. This approach is commonly used for configuration tables or tables necessitating frequent joins with minimal data updates. By consolidating joint queries and transactions between two tables onto a single node, a broadcast table enhances performance.
Single Table: A standalone table that doesn’t require sharding, often suitable for managing small datasets. As it remains unpartitioned, a single table seamlessly integrates with MySQL for streamlined usage.
Globally Unique Numerical Sequence
This is akin to AUTO_INCREMENT. The auto-increment sequence furnishes a globally unique numerical ID service, guaranteeing the global uniqueness of data entries like unique keys and primary keys within a distributed environment.
Command Passthrough
SQL statements can be directed to a designated target shard via annotation (hint), enhancing both performance and operational flexibility.
Secondary Partition
ADSQL further supports table partitioning alongside sharding, introducing an additional layer of logical partitioning (referred to as secondary partitioning) to horizontally sharded tables. This capability helps balance data distribution and access, enhances query efficiency, and proves beneficial for business scenarios requiring regular deletion of historical data, such as game match scores and IoT sensor data.
Read-write Separation (Read-only Slave)
ADSQL inherently supports read-write separation, where each slave within the master-slave architecture can operate in a read-only mode. When multiple slaves are configured, TProxy automatically directs read requests to low-load slaves to efficiently handle the read traffic of extensive applications. The read-write separation encompasses three schemes: read-only account, /*slave*/, and read-only instance.
Thread Pool
ADSQL includes a default thread pool feature, which incorporates optimized scheduling algorithms. These algorithms are designed to handle extreme scenarios where query and update requests are unevenly distributed across thread groups during periods of heavy system load. Moreover, these optimizations enhance computing resource utilization by minimizing unnecessary thread switching and expedite request processing by reducing the wait time of queued requests.
Premium Hardware Configuration
Leveraging PCI-E SSD technology, ADSQL boasts three times the IOPS of SATA, ensuring robust database accessibility. Its storage devices utilize the NvMe protocol, tailor-made for PCI-E SSDs, maximizing ADSQL’s performance potential. With a single shard supporting up to 245,509 queries per second (QPS), 384 GB memory, and 6 TB storage, ADSQL delivers exceptional performance.
Linear Scalability in Concurrency Performance
Within a distributed architecture, as the number of ADSQL shards increases, each shard takes on a portion of distributed tasks, enabling linear scalability in database performance.
RocksDB Support
RocksDB, an open-source storage engine utilized by numerous companies, including Facebook, is built on the LSM tree structure. It boasts a remarkable data compression rate exceeding 60% compared to the InnoDB engine. For instance, 160 GB of InnoDB-compressed data only occupies 62 GB of storage space in RocksDB. This translates to just 1 GB of space required for 10 million rows of data, significantly reducing database usage costs. Additionally, RocksDB exhibits superior write performance, exceeding five times that of InnoDB in high-concurrency scenarios. As a result, RocksDB is particularly well-suited for space-sensitive business applications characterized by a higher volume of writes than reads, such as IoT and logging.
Distributed High Availability
A distributed architecture inherently offers exceptional availability, as the failure of a single node does not impact the overall availability of the cluster. Building upon this foundation, ADSQL enhances availability by implementing master-slave redundancy for each shard, ensuring uninterrupted high availability for your business operations.
Three Data Replication Methods
ADSQL provides support for strong synchronous (non-downgradable), strong synchronous (downgradable), and asynchronous replication. Utilizing Alto’s proprietary multi-threaded asynchronous replication (MAR) technology ensures robust performance for strong synchronous data replication. This ensures master-slave data consistency, as the performance of strong synchronous replication is effectively equivalent to that of asynchronous replication.
Transparent Failove
Continuous monitoring of all shards and underlying physical devices is conducted around the clock. In the event of a failure, TProxy swiftly switches from the failing node to a functioning one. During master-slave switchover, the VIP remains unchanged, enabling seamless hot backup and switching at the application layer without any necessary modifications. This ensures that any failover has no impact on your business operations whatsoever.
Automatic Node Failure Recovery
In the event of a physical node failure carrying shards, the scheduling system initiates automatic retries to recover the node. If the original node cannot be restored, it automatically requests new resources within 30 minutes. It then rebuilds the node from backups and seamlessly integrates it into the cluster to ensure sustained functionality of master-slave switchover for the shards over the long term.
Cross-AZ Deployment
The master and slave can be deployed across different availability zones within the same city, enabling real-time data replication through Alto Cloud Direct Connect. The local end serves as the master, while the remote end functions as the slave. External data access initially targets the local instance. However, if it encounters failure or becomes inaccessible, the remote instance is automatically accessed. This capability equips ADSQL with cross-availability zone disaster recovery capability, ensuring transparency in master-slave switchover for end users and mitigating operational risks associated with single-IDC deployments.
Finance-grade Disaster Recovery
ADSQL offers deployment architectures like 1-region-2-DC or 2-region-3-DC. In case of a failure, the database can be swiftly restored to its normal state within minutes.
Compliant with Chinese, International, and Industry-specific Standards
ADSQL has obtained numerous Chinese and international certifications for Cloud Database, including ISO22301, ISO27001, ISO20000, ISO9001, Trusted Cloud, Information Security Protection Level (Level 3 or above), and CSA STAR, among others.
Certain functions within ADSQL have been developed and designed in accordance with standards such as GBT 20273-2006: Information Security Technology – Security Techniques Requirement for Database Management System and JRT 0072-2012: Testing and Evaluation Guide for Classified Protection of Information System of Financial Industry. This ensures that ADSQL meets even the most stringent database security requirements, providing users with complete peace of mind regarding their data security.
Security at the Database Kernel Level
ADSQL incorporates a comprehensive range of security optimizations to effectively mitigate security risks, including SQL injections, path traversals, unauthorized privilege escalation, system table overwriting attacks, and plugin attacks. Additionally, it features an embedded firewall within the database cluster, enhancing the overall security posture of your business operations and usage.
VPC Support
ADSQL permits the deployment of database instances within a Virtual Private Cloud (VPC), facilitating the isolation of your instances. You can seamlessly connect these instances to your existing IDC using either IPsec-encrypted VPN or Direct Connect connections.
Database Firewall and Security Protection
ADSQL offers default multiple layers of security protection for each database, effectively safeguarding against diverse attacking traffic. Even during DDoS attacks on internet access-enabled database instances, these layers ensure business continuity. Additionally, ADSQL efficiently counters database attacks like SQL injections and brute force attacks, significantly minimizing business interruption and associated losses.
Granular Permission Control
By default, accessing the database as the root account is disabled to mitigate security risks. ADSQL offers access controls at the object levels, including tables, functions, and storage processes. This ensures that only designated accounts can access permitted resources, thereby maintaining risks within a manageable scope.
Physical Exclusivity Support
All resources within a dedicated physical cluster (device) in the public cloud are exclusively available for deploying database instances. This means that you have sole access to the physical devices without sharing them with other tenants. Dedicated clusters effectively fulfill requirements for resource exclusivity, physical security, and compliance with industry regulations. Following the purchase of a dedicated cluster, database instances can be easily created with various customized specifications via the Cloud Console.
High Storage Reliability
ADSQL offers master-slave online data storage to uphold data security. Automatic daily backups are conducted, allowing for rollback to any point within the past three days based on backup files. Additionally, daily cold data is backed up in multiple copies, simplifying data recovery in case of a disaster. With ADSQL, you can confidently store your data without concerns of data loss.
Comprehensive Daily Monitoring
ADSQL offers instance performance monitoring at a granularity of 60 seconds, encompassing over 70 key metrics like connection access, database load, query cache, and storage engine. This comprehensive monitoring enables continuous assessment of database operational conditions. Additionally, 15 core database performance metrics support real-time monitoring with precision to the second, facilitating instance health tracking and swift detection of performance issues. Custom alarming capabilities for resource thresholds generate early warnings on database operational issues, alerting OPS personnel promptly for rapid response and prevention of data loss. Furthermore, slow log analysis reports and comprehensive SQL running reports are available for download, aiding in understanding factors influencing database performance.
Custom Alarming
Custom alarms for resource thresholds ensure proactive management of database operational issues. OPS personnel receive notifications promptly upon issue identification, enabling swift response and resolution.
Performance Analysis Logs
You can download slow log analysis reports and comprehensive SQL running reports to gain insights into the factors impacting database performance.
Elastic Scalability
Currently, a single shard can accommodate up to 6 TB of storage. If additional performance or capacity is needed, scaling up can be effortlessly achieved with just a few clicks in the console. There’s no need to fret over data migration, balancing, or route switching within the distributed system during the migration process—IP addresses remain unchanged, and service interruption lasts just a few seconds upon automatic switch. Ensuring your business has a reconnection mechanism is the only requirement.
Auto Rebalance
When shards are added or upgraded, data will be automatically migrated to ensure even data distribution across nodes. Your business will not be interrupted during the migration process, and only a few shards will become read-only for just seconds. This helps improve the efficiency of large-scale businesses and minimize the risks with OPS incidents.
Easy Management of Massive Databases
ADSQL management is streamlined through both command line and console interfaces, simplifying the management of extensive database instances into a few straightforward clicks online. This significantly reduces the workload for OPS personnel.
Multiple Network Access Methods
ADSQL offers support for access from both Virtual Private Cloud (VPC) and basic network configurations. External access can also be configured, allowing it to be accessed from IDCs, private clouds, or clouds provided by other vendors. This ensures that ADSQL meets your database access requirements across diverse environments.
API Support
ADSQL boasts a comprehensive API system, enabling seamless integration into your internal monitoring and OPS systems. Through APIs, you can create a business OPS system that automatically responds to your business needs. For detailed information, refer to the API documentation.
Compute
Edge Computing
Container
Distributed Cloud
Microservice
Serverless
Essential Storage Service
Data Migration
Data Process and Analysis
Relational Database
NoSQL Database
Enterprise Distributed DBMS
Database SaaS Tool
Networking
CDN and Edge platform
Network Security
Data Security
Application Security
Endpoint Security
Business Security
Security Services
Security Management
Domains & Websites
Office Collaboration
Enterprise Applications
Data Analysis
Image Recognition
Face Recognition
Voice Technology
AI Platform Service
Natural Language Processing
Optical Character Recognition
Internet of Things
Message Queue
Middleware
Communication
Interactive Video Services
Stream Services
Media On-Demand
Media Process Services
Media SDK
Cloud Real-time Rendering
Game Services
Game Video Service
Education Services
Blockchain Service
Building Services
Cloud Resource Management
Management and Audit Tools
Developer Tools
Monitor and Operation
Education