Today, large enterprise application platforms are transforming into microservice architecture. In  microservice architecture, the relationship between underlying platforms, such as applications and databases, will be refactored. As a new generation of distributed database, SequoiaDB provides an extremely important boost for the cloud infrastructure upgrade of many large financial customers.

As a new generation of distributed database, SequoiaDB's architecture and functional characteristics need to ensure that it embraces microservice and cloud computing frameworks based on fully compatibility with traditional databases. Therefore, distributed databases must be fully compatible with traditional technologies for distributed transactions and ACID. At the same time, the new generation of distributed databases must support horizontal scalability, resource isolation, multi-tenancy, configurable consistency, multi-model (supports various SQL protocols), and configurable disaster recovery policy within the cluster.

The capacity bottleneck of traditional single-point database is only one of the problems solved by distributed databases. The applications are no longer built in the “stacked” middleware plus database model, but they are built using thousands or even tens of thousands of microservice to build a complex mesh model. Therefore, distributed databases must have the following capabilities to meet the elastic scalability, high concurrency, high throughput, and flexible agility requirements of upper-layer applications.

Driven by these technical requirements, the distributed databases core technical capabilities are divided into two aspects.

One is the compatibility with traditional technologies, including:

·  Complete ACID support, transaction and consistency guarantee

·  Full support for SQL and compatible with traditional databases such as MySQL/PostgreSQL

On the other hand, its technological innovation, including:

· Distributed and scalable architecture, responding to changes in data volume, and enabling flexible expansion of storage and computing layers

· Support multiple type of data management and multiple modes of  access to the interface

· HTAP (Hybrid Transactional/Analytical Processing) capabilities,  physical isolation of data without interference.

As an enterprise-class distributed relational database, SequoiaDB's distributed database architecture and cloud-oriented microservice product model have helped a number of large-scale financial customers, including Minsheng Bank and Evergrowing Bank. At present, the single cluster bank production system of SequoiaDB has 135 physical nodes with the maximum storage capacity of 2.1 PB, and the maximum number of records of 131.8 billion.

Data storage resource pooling

The data storage engine of SequoiaDB uses a native distributed architecture, and data is completely scattered between distributed nodes, achieving automated data distribution and management, and data can be flexibly expanded as needed.

SequoiaDB uses sharding technology to provide a scale-out mechanism for system. The sharding process is completely transparent to applications. And it addresses the limitation of single-server hardware resources (such as memory, CPU, disk I/O) and does not increase the complexity of application development. SequoiaDB can easily implement Petabyte-level data management through the natively distributed architecture. Currently, the production environment supports more than 1500 node clusters.

SequoiaDB storage engine also implements multi-model data management, supporting unstructured, structured and semi-structured data coverage and unified management. SequoiaDB's multi-model engine design makes the database platform scenario more diverse, and can also meet the unified management, operation and maintenance requirement for diverse business data with cloud data architecture.

In a large cluster, SequoiaDB provides multi-dimensional, multi-level logical and physical isolation capabilities. In a typical data resource pooling data service platform (DBaaS, DataBase As A Service), SequoiaDB can serve hundreds of applications with different SLA, priorities, business characteristics, and data volume. It can also ensure logical and physical isolation of data between applications.

SequoiaDB provides the "Domain" feature to isolate the entire cluster at the physical device level, ensuring that different tables, table-spaces, and database instances are located on independent hardware devices, ensuring high-priority online transaction applications to be isolated from background statistical analysis and does not interfere with each other.

Database Instantiation

SequoiaDB supports database service instantiation.

For the microservice application architecture, users can create hundreds or thousands of different relational database instances in the same cluster. Database instances are 100% compatible with traditional relational databases, and the underlying data is logically completely independent. Each instance has its own independent rights management, data management, and it can be deployed in private hardware environment or shared device.

Currently, SequoiaDB supports users to create MySQL, PostgreSQL and SparkSQL instances. It also provides JSON, S3 object storage and Posix file system instances to fully meet the user's needs for structured, semi-structured, and unstructured data.

From the perspective of application developers and DBAs, the relational database instances provided by SequoiaDB are fully compatible with traditional MySQL, PostgreSQL and SparkSQL. For example, in the MySQL instance of SequoiaDB, all its CRUD syntax, views, triggers, transactions, and even access plans are consistent with traditional MySQL.

As a distributed database, the SQL instance user of SequoiaDB does not need to care whether the underlying data is scattered among one or more devices. Users can simply create a partition table, write hundreds of millions of records to it. The data will be automatically dispersed in different physical devices. The application does not need to pay attention to the sub-database table at all, and the database automatically provides abilities of distributed business and distributed access.

Disaster recovery and data security

In the dual-active disaster recovery system, the main production end database and the standby end database are simultaneously running online, and are in a readable and writable state. In the bank's transaction system, the dual-active disaster recovery capability not only ensures that the data are not lost, but also can resume service in a short time when one DC is down. Under normal circumstances, both data centers can provide read and write services for business at the same time. When a central server is down, all front-end applications can immediately switch to the alive data center. Through SequoiaDB’s mechanism of disaster recovery and high availability, the server failure in the data center can ensure that both the RTO and the RPO are zero, and the entire data center or network failure in the same city can also achieve the second-level RTO, RPO=0.

About SequoiaDB

SequoiaDB is the leading distributed database company in China. Since the establishment in 2011, SequoiaDB has been dedicated to the research and development of the new generation distributed database management system. In 2017 and 2018, SequoiaDB has been listed in Gartner’s database report, and is the first Chinese DBMS vendor listed.

The main products of SequoiaDB includes distributed relational database and SequoiaCM enterprise content management software. The application scenarios includes: distributed online transaction, data mid-end, distributed content management and etc.

SequoiaDB now has more than 1000 enterprise customers and community users. SequoiaDB has been deployed on core systems of more than 50 major financial institutions, like banking, insurance and security clients of Fortune-500 Level.