- Releases Notes and Announcements
- Product Introduction
- Purchase Guide
- Getting Started
- Operation Guide
- SQL Developer Guide
- Python Developer Guide
- ETL Developer Guide
- FAQ
- Contact Us
- Releases Notes and Announcements
- Product Introduction
- Purchase Guide
- Getting Started
- Operation Guide
- SQL Developer Guide
- Python Developer Guide
- ETL Developer Guide
- FAQ
- Contact Us
The table below lists some of the commonly used terms related to SQL jobs.
Term | Description |
Stream computing | Stream computing is the computing of stream data. It reads data in stream form from one or more data sources, efficiently computes the continuous data streams using multiple operators of the engine, and outputs the results to different sinks such as message queues, databases, data warehouses, and storage services. |
Data source | The source that continuously generates data for stream computing, such as CKafka. |
Data sink | The destination of the results of stream computing, such as CKafka, TencentDB for MySQL, and TencentDB for PostgreSQL. |
Schema | The structure information of a table, such as column headings and types. In the context of PostgreSQL, a schema is smaller than a database and larger than a table. It can be seen as a namespace inside a database. |
Time mode | The time mode determines how the system obtains timestamp information when processing data. Currently, three time modes are supported, namely Event Time, Processing Time, and Source Time |
Event Time | In the Event Time mode, timestamp information is offered by an input field. You can use the WATERMARK FOR statement to specify this field and enable the Event Time mode. This mode is suitable for scenarios where the data source offers precise timestamp information. |
Watermark | A watermark is a time point. All data before this time point have been properly processed. Watermarks are automatically generated. You can use the WATERMARK FOR BOUNDED statement to specify the maximum error tolerance. |
Processing Time | In the Processing Time mode, the system automatically generates timestamps ( PROCTIME) and adds them to the data source (PROCTIME is invisible to SELECT *. You need to specify it explicitly). In this mode, a timestamp is the time each data record is processed. Because such data has some uncertainty, this mode is suitable for scenarios that do not have high requirements on precision. |
Source Time | In the Source Time mode, you can use the timestamp in the metadata of each Kafka record as the timestamp ( SOURCETIME) for stream computing (SOURCETIME is invisible to SELECT *. You need to specify it explicitly). In cases where the input data does not include a timestamp field, this mode eliminates the uncertainty of the Processing Time mode. |
Time window | A time window specifies multiple time ranges and the relationships among them, for example, whether they can overlap or whether the window length is fixed. Currently, three types of time windows are supported, namely TUMBLE, HOP, and SESSION. For details, see Time Window Functions. |
TencentDB for MySQL | TencentDB is a database management service featuring high performance, availability, and scalability. It helps you easily deploy and use MySQL, PostgreSQL, and other databases on the cloud. |
CKafka | CKafka is a high-throughput, highly scalable distributed messaging system that is fully compatible with open-source Apache Kafka API 0.10. Currently, Stream Compute Service supports inputs and outputs in CSV and JSON formats. |
Tuple (Append) stream | A tuple (append) can store stream data that does not contain a primary key. You can append data to a tuple continuously. Already sent data is not affected. Currently, append streams are supported by all data sources and sinks. |
Upsert stream | Upsert (update or insert) streams are generated by queries such as DISTINCT, GROUP BY without a time window, and JOIN without a time range. A primary key is defined for upsert streams. If a new data record has the same primary key as an existing record, the existing data will be updated. Otherwise, a new record will be added. This ensures that existing data is up-to-date. |
DDL statements | Data Definition Language (DDL) is a subset of SQL consisting of CREATE statements. You can use DDL statements to define a table, a view, and a user-defined function (UDF). |
DML statement | Data Manipulation Language (DML) is a subset of SQL consisting of INSERT and SELECT statements. You can use DML statements to select, change, filter, and insert tables and views. |
Yes
No
Was this page helpful?