StarRocks

Overview

• StarRocks Load node supports writing data to the StarRocks database.
• Two modes are supported for sink to StarRocks: Single-sink for specify fixed database name and table name to sink. Multi-sink for custom database name and table name according to src format, which suitable for scenarios such as multi-table writing or whole database synchronization.
• This document describes how to set up a StarRocks Load node to sink to StarRocks.

Supported Version

Load Node	StarRocks version
StarRocks	2.0+

Dependencies

In order to set up the StarRocks Load node, the dependency information needed to use a build automation tool such as Maven or SBT is provided below.

Maven dependency

<dependency>
    <groupId>org.apache.inlong</groupId>
    <artifactId>sort-connector-starrocks</artifactId>
    <version>1.7.0</version>
</dependency>

## Prepare
### Create MySql Extract table
- For Single-sink: Create a table `cdc.cdc_mysql_source` in the MySQL database. The command is as follows:
```sql
[root@fe001 ~]# mysql -u root -h localhost -P 3306 -p123456
mysql> use cdc;
Database changed
mysql> CREATE TABLE `cdc_mysql_source` (
       `id` int(11) NOT NULL AUTO_INCREMENT,
       `name` varchar(64) DEFAULT NULL,
       `dr` tinyint(3) DEFAULT 0,
       PRIMARY KEY (`id`)
       );
Query OK, 0 rows affected (0.02 sec)

mysql> insert into cdc_mysql_source values(1, 'zhangsan', 0),(2, 'lisi', 0),(3, 'wangwu', 0);
Query OK, 3 rows affected (0.01 sec)
Records: 3  Duplicates: 0  Warnings: 0

mysql> select * from cdc_mysql_source;
+----+----------+----+
| id | name     | dr |
+----+----------+----+
|  1 | zhangsan |  0 |
|  2 | lisi     |  0 |
|  3 | wangwu   |  0 |
+----+----------+----+
3 rows in set (0.07 sec)

• For Multi-sink: Create tables user_db.user_id_name、user_db.user_id_score in the MySQL database. The command is as follows:

[root@fe001 ~]# mysql -u root -h localhost -P 3306 -p123456
mysql> use user_db;
Database changed
mysql> CREATE TABLE `user_id_name` (
       `id` int(11) NOT NULL AUTO_INCREMENT,
       `name` varchar(64) DEFAULT NULL
       PRIMARY KEY (`id`)
       );
Query OK, 0 rows affected (0.02 sec)

mysql> CREATE TABLE `user_id_score` (
       `id` int(11) NOT NULL AUTO_INCREMENT,
       `score` double default 0,
       PRIMARY KEY (`id`)
       );
Query OK, 0 rows affected (0.02 sec)

mysql> insert into user_id_name values(1001, 'lily'),(1002, 'tom'),(1003, 'alan');
Query OK, 3 rows affected (0.01 sec)
Records: 3  Duplicates: 0  Warnings: 0 

mysql> insert into user_id_score values(1001, 99),(1002, 96),(1003, 98);
Query OK, 3 rows affected (0.01 sec)
Records: 3  Duplicates: 0  Warnings: 0 

mysql> select * from user_id_name;
+------+--------+
|  id  | name   |
+------+--------+
| 1001 | lily   |
| 1002 | tom    |
| 1003 | alan   |
+----+----------+
3 rows in set (0.07 sec)    

mysql> select * from user_id_score;
+------+------+
|  id  | name |
+------+------+
| 1001 | 99   |
| 1002 | 96   |
| 1003 | 98   |
+----+--------+
3 rows in set (0.07 sec)  

Create StarRocks Load table

• For Single-sink: Create a table cdc.cdc_starrocks_sink in the StarRocks database. The command is as follows:

[root@fe001 ~]# mysql -u username -h localhost -P 9030 -p password
mysql> use cdc;
Reading table information for completion of table and column names
You can turn off this feature to get a quicker startup with -A
Database changed

mysql> CREATE TABLE `cdc_starrocks_sink` (
       `id` int(11) NOT NULL COMMENT "user id",
       `name` varchar(50) NOT NULL COMMENT "user name",
       `dr` tinyint(4) NULL COMMENT "delete tag"
       ) ENGINE=OLAP
       PRIMARY KEY(`id`)
       COMMENT "OLAP"
       DISTRIBUTED BY HASH(`id`) BUCKETS 1
       PROPERTIES (
       "replication_allocation" = "tag.location.default: 1"
       );
Query OK, 0 rows affected (0.06 sec)

• For Multi-sink: Create tables user_db.starrocks_user_id_name、user_db.starrocks_user_id_score in the StarRocks database. The command is as follows:

[root@fe001 ~]# mysql -u username -h localhost -P 9030 -p password
mysql> use user_db;
Reading table information for completion of table and column names
You can turn off this feature to get a quicker startup with -A
Database changed

mysql> CREATE TABLE `starrocks_user_id_name` (
       `id` int(11) NOT NULL COMMENT "用户id",
       `name` varchar(50) NOT NULL COMMENT "昵称"
       ) ENGINE=OLAP
       PRIMARY KEY(`id`)
       COMMENT "OLAP"
       DISTRIBUTED BY HASH(`id`) BUCKETS 1
       PROPERTIES (
       "replication_allocation" = "tag.location.default: 1"
       );
Query OK, 0 rows affected (0.06 sec)

mysql> CREATE TABLE `starrocks_user_id_score` (
       `id` int(11) NOT NULL COMMENT "用户id",
       `score` double default 0
       ) ENGINE=OLAP
       PRIMARY KEY(`id`)
       COMMENT "OLAP"
       DISTRIBUTED BY HASH(`id`) BUCKETS 1
       PROPERTIES (
       "replication_allocation" = "tag.location.default: 1"
       );
Query OK, 0 rows affected (0.06 sec)

How to create a StarRocks Load Node

Usage for SQL API

• For Single-sink: StarRocks load

[root@tasknode001 flink-1.13.5]# ./bin/sql-client.sh -l ./opt/connectors/mysql-cdc-inlong/ -l ./opt/connectors/starrocks/
Flink SQL> SET 'execution.checkpointing.interval' = '3s';
[INFO] Session property has been set.

Flink SQL> SET 'table.dynamic-table-options.enabled' = 'true';
[INFO] Session property has been set.

Flink SQL> CREATE TABLE cdc_mysql_source (
    >   id int
    >   ,name VARCHAR
    >   ,dr TINYINT
    >   ,PRIMARY KEY (id) NOT ENFORCED
    > ) WITH (
    >  'connector' = 'mysql-cdc-inlong',
    >  'hostname' = 'localhost',
    >  'port' = '3306',
    >  'username' = 'root',
    >  'password' = '123456',
    >  'database-name' = 'cdc',
    >  'table-name' = 'cdc_mysql_source'
    > );
[INFO] Execute statement succeed.

Flink SQL> CREATE TABLE cdc_starrocks_sink (
    > id INT,
    > name STRING,
    > dr TINYINT
    > ) WITH (
    >  'connector' = 'starrocks-inlong',
    >  'fenodes' = 'localhost:8030',
    >  'table.identifier' = 'cdc.cdc_starrocks_sink',
    >  'username' = 'username',
    >  'password' = 'password',
    >  'sink.properties.format' = 'json',
    >  'sink.properties.strip_outer_array' = 'true'
    > );
[INFO] Execute statement succeed.

Flink SQL> insert into cdc_starrocks_sink select * from cdc_mysql_source /*+ OPTIONS('server-id'='5402') */;
[INFO] Submitting SQL update statement to the cluster...
[INFO] SQL update statement has been successfully submitted to the cluster:
Job ID: 5f89691571d7b3f3ca446589e3d0c3d3

• For Single-sink: StarRocks load

./bin/sql-client.sh -l ./opt/connectors/mysql-cdc-inlong/ -l ./opt/connectors/starrocks/
Flink SQL> SET 'execution.checkpointing.interval' = '3s';
[INFO] Session property has been set.

Flink SQL> SET 'table.dynamic-table-options.enabled' = 'true';
[INFO] Session property has been set.

Flink SQL> CREATE TABLE cdc_mysql_source (
    >   id int
    >   ,name VARCHAR
    >   ,dr TINYINT
    >   ,PRIMARY KEY (id) NOT ENFORCED
    > ) WITH (
    >  'connector' = 'mysql-cdc-inlong',
    >  'hostname' = 'localhost',
    >  'port' = '3306',
    >  'username' = 'root',
    >  'password' = '123456',
    >  'database-name' = 'test',
    >  'table-name' = 'cdc_mysql_source'
    > );
[INFO] Execute statement succeed.

Flink SQL> CREATE TABLE cdc_starrocks_sink (
    > id INT,
    > name STRING,
    > dr TINYINT
    > ) WITH (
    >  'connector' = 'starrocks-inlong',
    >  'fenodes' = 'localhost:8030',
    >  'username' = 'username',
    >  'password' = 'password',
    >  'sink.multiple.enable' = 'true',
    >  'sink.multiple.format' = 'canal-json',
    >  'sink.multiple.database-pattern' = '${database}',
    >  'sink.multiple.table-pattern' = 'starrocks_${table}'
    > );
[INFO] Execute statement succeed.

Flink SQL> insert into cdc_starrocks_sink select * from cdc_mysql_source /*+ OPTIONS('server-id'='5402') */;
[INFO] Submitting SQL update statement to the cluster...
[INFO] SQL update statement has been successfully submitted to the cluster:
Job ID: 30feaa0ede92h6b6e25ea0cfda26df5e

Usage for InLong Dashboard

TODO: It will be supported in the future.

Usage for InLong Manager Client

TODO: It will be supported in the future.

StarRocks Load Node Options

Option	Required	Default	Type	Description
connector	required	(none)	string	Specify which connector to use, valid values are: starrocks-inlong
jdbc-url	required	(none)	string	this will be used to execute queries in starrocks.
load-url	required	(none)	string	fe_ip:http_port;fe_ip:http_port separated with ';', which would be used to do the batch sinking.
database-name	required	(none)	string	starrocks database name
table-name	required	(none)	string	starrocks table name
username	required	(none)	string	starrocks connecting username
password	required	(none)	string	starrocks connecting password
sink.semantic	optional	at-least-once	string	at-least-once or exactly-once(flush at checkpoint only and options like sink.buffer-flush.* won't work either).
sink.version	optional	AUTO	string	The version of implementaion for sink exactly-once. Only availible for connector 1.2.4+. If V2, use StarRocks' stream load transaction interface which requires StarRocks 2.4+. If V1, use stream load non-transaction interface. If AUTO, connector will choose the stream load transaction interface automatically if the StarRocks supports the feature, otherwise choose non-transaction interface.
sink.buffer-flush.max-bytes	optional	94371840(90M)	string	the max batching size of the serialized data, range: [64MB, 10GB].
sink.buffer-flush.max-rows	optional	500000	string	the max batching rows, range: [64,000, 5000,000].
sink.buffer-flush.interval-ms	optional	300000	string	the flushing time interval, range: [1000ms, 3600000ms].
sink.max-retries	optional	3	string	max retry times of the stream load request, range: [0, 10].
sink.connect.timeout-ms	optional	1000	string	Timeout in millisecond for connecting to the load-url, range: [100, 60000].
sink.properties.format	optional	CSV	string	The file format of data loaded into starrocks. Valid values: CSV and JSON. Default value: CSV.
sink.properties.*	optional	NONE	string	the stream load properties like 'sink.properties.columns' = 'k1, k2, k3',details in STREAM LOAD. Since 2.4, the flink-connector-starrocks supports partial updates for Primary Key model.
sink.properties.ignore_json_size	optional	false	string	ignore the batching size (100MB) of json data
sink.multiple.enable	optional	false	boolean	Determine whether to support multiple sink writing, default is false. when sink.multiple.enable is true, need sink.multiple.format、sink.multiple.database-pattern、sink.multiple.table-pattern be correctly set.
sink.multiple.format	optional	(none)	string	The format of multiple sink, it represents the real format of the raw binary data. can be canal-json or debezium-json at present. See kafka -- Dynamic Topic Extraction for more details.
sink.multiple.database-pattern	optional	(none)	string	Extract database name from the raw binary data, this is only used in the multiple sink writing scenario.
sink.multiple.table-pattern	optional	(none)	string	Extract table name from the raw binary data, this is only used in the multiple sink writing scenario.
inlong.metric.labels	optional	(none)	String	Inlong metric label, format of value is groupId={groupId}&streamId={streamId}&nodeId={nodeId}.
sink.multiple.schema-update.policy	optional	(none)	string	If sink data fields do not match starrocks table, such as table does not exsit or string data is over length, starrocks server will throw an exception. When this option is THROW_WITH_STOP, the exception will be thrown up to flink framework, flink will restart task automatically, trying to resume the task. When this option is STOP_PARTIAL, starrocks connector will stop writing into this table, other tables are written normally. The exception will be logging but not thrown up. When this option is LOG_WITH_IGNORE, starrocks connector only log the error, not throw up. StarRocks connector will try to write to starrocks server again when receiving new source data.
dirty.ignore	optinal	(none)	boolean	When writing data into starrocks table, errors may be thrown by starrocks server as table does not exist or data is over length. When this option is true, and dirty.side-output.* properties are configed correctly, dirty data can be written to Amazon S3 or Tencent Colud COS storage. Dirty data metrics will also be collected automatically. When this option is false, only dirty data metrics will be collected, but dirty data will not be archived.
dirty.side-output.enable	optinal	(none)	boolean	When this option is ture and other options about S3 or COS is configed correctly, dirty data archiving will works. When false, dirty data archiving will not work.
dirty.side-output.connector	optinal	(none)	string	s3 or log are supported now. When log, starrocks connector only log the dirty data, not archive data. When s3, starrocks connector can write dirty data to S3 or COS.
dirty.side-output.s3.bucket	optinal	(none)	string	The bucket name of S3 or COS
dirty.side-output.s3.endpoint	optinal	(none)	string	The endpoint of S3 or COS
dirty.side-output.s3.key	optinal	(none)	string	The key of S3 or COS
dirty.side-output.s3.region	optinal	(none)	string	The region of S3 or COS
dirty.side-output.line-delimiter	optinal	(none)	string	The line delimiter of dirty data
dirty.side-output.field-delimiter	optinal	(none)	string	The field delimiter of dirty data
dirty.side-output.s3.secret-key-id	optinal	(none)	string	The secret key of S3 or COS
dirty.side-output.s3.access-key-id	optinal	(none)	string	The access key of S3 or COS
dirty.side-output.format	optinal	(none)	string	The format of dirty data archiving, supports json or csv
dirty.side-output.log-tag	optinal	(none)	string	The log tag of dirty data. StarRocks connector uses lags to distinguish which starrocks database and table the dirty data will be written to.
dirty.identifier	optinal	(none)	string	The file name of drity data which written to S3 or COS.
dirty.side-output.labels	optinal	(none)	string	Every dirty data line contains label and business data fields. Label is in front, and business data is at end.

Data Type Mapping

Flink type	StarRocks type
BOOLEAN	BOOLEAN
TINYINT	TINYINT
SMALLINT	SMALLINT
INTEGER	INTEGER
BIGINT	BIGINT
FLOAT	FLOAT
DOUBLE	DOUBLE
DECIMAL	DECIMAL
BINARY	INT
CHAR	JSON / STRING
VARCHAR	JSON / STRING
STRING	JSON / STRING
DATE	DATE
TIMESTAMP_WITHOUT_TIME_ZONE(N)	DATETIME
TIMESTAMP_WITH_LOCAL_TIME_ZONE(N)	DATETIME
ARRAY<T>	ARRAY<T>
MAP<KT,VT>	JSON / JSON STRING
ROW<arg T...>	JSON / JSON STRING

See flink-connector-starrocks for more details.