What are distributed databases?

Distributed databases include: 1. In the direction of the Internet of Things, time series database products meet the collection, storage and statistics of IoT data, such as InfluxDB, Kudu, kdb, and OpenTSDB; 2. In the direction of transaction relationships, Ant Financial Oceanbase , Tencent TDSQL.

Distributed database products on the market are divided into several categories:

1. Internet of Things direction: time series database products to meet the collection of IoT data , storage and statistics. Time series database products are also the ones that have the greatest impact on memory database products. For example: InfluxDB, Kudu, kdb, OpenTSDB;

2. Transaction relationship direction: replace traditional transaction relational database products Oracle/DB2 and other online transaction businesses that cannot meet the needs of massive throughput, massive concurrency, massive transactions, and massive storage. Scenes. For example: Ant Financial Oceanbase, Tencent TDSQL, Repu HotDB, ZTE GoldenDB, open source MyCAT, open source Cobar.

3. Analysis relationship direction: solve business scenarios of structured data storage and data analysis, such as: Greenplum, Vertical, Gbase8a, etc. However, this area has received a huge impact from KV analysis products;

4. KV analysis direction: Hadoop and Spark are the current cornerstones. Many domestic and foreign companies are doing secondary research and development based on them. In particular, it is compatible with SQL standard syntax and has catered to business scenarios and R&D personnel.

5. KV document direction: solve the unstructured data storage and data processing of online document types, such as MongoDB and SequoiaDB, but they are also trying their best to be compatible with SQL standard syntax.

6. HTAP: Transaction analysis hybrid distributed database product. From the perspective of technical principles, this is a direction without theoretical innovation support. It is just the inner wish of our technical staff, such as: domestic TiDB, foreign Spanner/F1 (no one knows what it looks like or what the experience is like).

Each route will have its own specific algorithm, specific architecture and product features. It is difficult to have a product that is compatible with all and has great performance.

The above is the detailed content of What are distributed databases?. For more information, please follow other related articles on the PHP Chinese website!