Note
This feature is supported starting from V4.6.2.1.
The flash application scenario is a practical application of logical tables, designed for dynamic table structures and multi-application SQL read/write: multiple small applications can share the same set of underlying physical tables, isolated from each other by App (i.e., Namespace), eliminating the need to create separate physical tables for each application.
This scenario uses logical tables with json-table-sdk (set use_namespace=True when creating the client). The usage is the same as for regular tables: first CREATE TABLE to create the table, then use SQL statements like INSERT and SELECT to read and write data. Rows are stored in JSON format within the database, and json-table-sdk automatically handles the conversion, so you generally do not need to directly process JSON paths or use the JSON_TABLE function.
Core capabilities
Capabilities |
Note |
|---|---|
| Multi-application isolation | Multiple apps share the underlying storage of a Collection but are logically isolated by Namespaces. |
| Creation and modification of virtual tables | You can create and delete virtual tables, add columns, and query table schemas and tables.
NoticeStarting from V4.6.2.1, you can create multiple tables in the same app (namespace). |
| SQL read/write | Execute INSERT, UPDATE, DELETE, and SELECT statements through the SDK. |
| JSON-based structured query | Use the JSON_TABLE function to restore the JSON row to a structured result (converted by json-table-sdk). |
| Search index acceleration | Logical data tables have built-in search indexes (such as idx_json(data_content)) to accelerate searches. |
| Partition pruning | Partitioning by namespace_id KEY allows a single app query to prune partitions. |
| Prewarm | When an app has not been accessed for a long time or has just gone live, the first query may be slow. You can call prewarm to load the app's data into the cache in advance to reduce first-access latency. |
References
Recommended reading order: First understand the capabilities on this page, then check Limitations and constraints to confirm the access boundaries, and then proceed to the connection and quick start to practice.
- For background information, object models of logical tables, and comparisons with hybrid search scenarios using Agents, see Logical table overview
- Limitations and constraints
- For connection and client parameters, see Connect to database by using json-table-sdk
- Quick start with flash application scenario
- Collection and app management
- Data operations
- Agent hybrid search scenario overview