Limitations¶
MatrixOne is compatible with a commonly used subset of MySQL, not all of MySQL 8.0. The following boundaries are important to understand before copying syntax from the upstream MySQL manual into production.
Limitations may be relaxed in later versions. Use error messages from the current instance and the release notes as the final authority.
LOAD DATA¶
The following parts of the complete MySQL LOAD DATA syntax are commonly unsupported or only partially supported in MatrixOne:
Item |
Description |
|---|---|
|
Uploading a client-local file to the server through a |
|
|
|
Loading into a specified partition is generally unsupported. |
|
Data is effectively handled as UTF-8. |
Arbitrary field or line delimiters |
Delimiters are supported, but some escapes, variable |
|
Column-expression transformations during loading are generally unsupported. |
User-variable columns |
Behaviors such as filtering through |
Recommended usage pattern:
LOAD DATA INFILE 'path/to/file'
[IGNORE]
INTO TABLE tbl
FIELDS TERMINATED BY ','
[IGNORE n LINES]
[(col_list)];
Summary of CSV conventions:
Field delimiter: Usually
,, though other non-special delimiters such as|can be used.Field quote:
".Line ending:
\nor\r\n.\Nrepresents NULL.Lines beginning with
#are commonly skipped as comments.IGNORE: When conversion fails, continue where possible and accumulate warnings instead of failing the entire batch. Exact behavior depends on the engine.
For large files, prefer loading from a Stage and server-side path instead of relying on client-side LOCAL.
Partitioning¶
Documentation and metadata may list partition type enums such as
KEY/HASH.Types such as
RANGE/LISTare generally still unavailable. Related description columns inmo_table_partitionsmay be empty.Do not plan capacity around the full set of MySQL partitioning strategies. First validate
CREATE TABLE ... PARTITION BYon the target instance.
JDBC and driver capabilities¶
Official JDBC and other drivers use a supported-feature list for MatrixOne. Common CRUD, prepared-statement, and batch operations are supported, while some proprietary or uncommon MySQL Connector APIs may not be implemented. During integration, refer to Developer and the driver release notes. Do not assume behavior is 100% identical to MySQL Connector/J.
Other common differences¶
Area |
Description |
|---|---|
Reserved keywords |
They differ slightly from MySQL. Reserved words in qualified names should still be enclosed in backticks. See Structure and Syntax. |
System tables |
|
Vector columns |
They cannot be primary or unique keys, and dimensions and element types are constrained. See Data Types. |
Stored procedures / triggers / event scheduler |
Do not assume they are available. Refer to the syntax actually supported by the instance. |
Character sets |
Product workflows primarily use |