Updated on 2024-11-08 GMT+08:00

Redis Result Table

Function

DLI outputs the Flink job output data to Redis. Redis is a key-value storage system that supports multiple types of data structures. It can be used in scenarios such as caching, event publish/subscribe, and high-speed queuing. Redis supports direct read/write of strings, hashes, lists, queues, and sets. Redis works with in-memory datasets and provides persistence. For more information about Redis, visit https://redis.io/.

Prerequisites

  • An enhanced datasource connection with Redis has been established, so that you can configure security group rules as required.
  • In Flink cross-source development scenarios, there is a risk of password leakage if datasource authentication information is directly configured. You are advised to use the datasource authentication provided by DLI.

    For details about datasource authentication, see Introduction to Datasource Authentication.

Precautions

  • When creating a Flink OpenSource SQL job, you need to set Flink Version to 1.12 on the Running Parameters tab of the job editing page, select Save Job Log, and set the OBS bucket for saving job logs.
  • If the Redis key field is not defined in the statement for creating the Redis result table, the generated UUID is used as the key.
  • To specify a key in Redis, you need to define a primary key in the Redis result table of Flink. The value of the primary key is the Redis key.
  • If the primary key defined for the Redis result table, it cannot be a composite primary key and only can be one field.
  • Constraints on schema-syntax:
    • If schema-syntax is map or array, there can be only one non-primary key and it must be of the same map or array type.
    • If schema-syntax is fields-scores, the number of non-primary keys must be an even number, and the second key of every two keys except the primary key must be of the double type. The double value is the score of the previous key. The following is an example:
      CREATE TABLE redisSink (
        order_id string,
        order_channel string,
        order_time double,
        pay_amount STRING,
        real_pay double,
        pay_time string,
        user_id double,
        user_name string,
        area_id double,
        primary key (order_id) not enforced
      ) WITH (
        'connector' = 'redis',
        'host' = 'RedisIP',
        'password' = 'RedisPassword',
        'data-type' = 'sorted-set',
        'deploy-mode' = 'master-replica',
        'schema-syntax' = 'fields-scores'
      );
  • Restrictions on data-type:
    • If data-type is string, only one non-primary key field is allowed.
    • If data-type is sorted-set and schema-syntax is fields or array, default-score is used as the score.
    • If data-type is sorted-set and schema-syntax is map, there can be only one non-primary key in addition to the primary key and the non-primary key must be of the map type. The map values of the non-primary key must be of the double type, indicating the score. The keys in the map are the values in the Redis set.
    • If data-type is sorted-set and schema-syntax is array-scores, only two non-primary keys are allowed and must be of the array type.
      The first key indicates values in the Redis set. The second key is of the array<double> type, indicating index scores. The following is an example:
      CREATE TABLE redisSink (
        order_id string,
        arrayField Array<String>,
        arrayScore array<double>,
        primary key (order_id) not enforced
      ) WITH (
        'connector' = 'redis',
        'host' = 'RedisIP',
        'password' = 'RedisPassword',
        'data-type' = 'sorted-set',
        "default-score" = '3',
        'deploy-mode' = 'master-replica',
        'schema-syntax' = 'array-scores'
      );

Syntax

1
2
3
4
5
6
7
8
9
create table dwsSink (
  attr_name attr_type 
  (',' attr_name attr_type)* 
  (','PRIMARY KEY (attr_name) NOT ENFORCED)
)
with (
  'connector' = 'redis',
  'host' = ''
);

Parameters

Table 1 Parameter description

Parameter

Mandatory

Default Value

Data Type

Description

connector

Yes

None

String

Connector to be used. Set this parameter to redis.

host

Yes

None

String

Redis connector address.

port

No

6379

Integer

Redis connector port.

password

No

None

String

Redis authentication password.

namespace

No

None

String

Redis key namespace.

For example, if the value is set to "person" and the key is "jack", the value in the Redis is person:jack.

delimiter

No

:

String

Delimiter between the Redis key and namespace.

data-type

No

hash

String

Redis data type. Available values are as follows:

  • hash
  • list
  • set
  • sorted-set
  • string

For details about the constraints, see Constraints on data-type.

schema-syntax

No

fields

String

Redis schema semantics. Available values are as follows:

  • fields: applicable to all data types. This value indicates that multiple fields can be set and the value of each field is read when data is written.
  • fields-scores: applicable to sorted-set data, indicating that each field is read as an independent score.
  • array: applicable to list, set, and sorted-set data.
  • array-scores: applicable to sorted-set data.
  • map: applicable to hash and sorted-set data.

For details about the constraints, see Constraints on schema-syntax.

deploy-mode

No

standalone

String

Deployment mode of the Redis cluster. The value can be standalone, master-replica, or cluster. The default value is standalone.

For details about the setting, see the instance type description of the Redis cluster.

retry-count

No

5

Integer

Number of attempts to connect to the Redis cluster.

connection-timeout-millis

No

10000

Integer

Maximum timeout for connecting to the Redis cluster.

commands-timeout-millis

No

2000

Integer

Maximum time for waiting for a completion response.

rebalancing-timeout-millis

No

15000

Integer

Sleep time when the Redis cluster fails.

default-score

No

0

Double

Default score when data-type is sorted-set.

ignore-retraction

No

false

Boolean

Whether to ignore Retract messages.

skip-null-values

No

true

Boolean

Whether null values will be skipped. If this parameter is false, null will be assigned for null values.

pwd_auth_name

No

None

String

Name of datasource authentication of the password type created on DLI.

If datasource authentication is used, you do not need to set the username and password for jobs.

key-ttl-mode

No

no-ttl

String

Whether the Redis sink TTL function will be enabled. The value can be no-ttl, expire-msec, expire-at-date or expire-at-timestamp.

  • no-ttl: No expiration time is set.
  • expire-msec: validity period of the key. The parameter is a long string, in milliseconds.
  • expire-at-date: Date and time when the key expires. The value is in UTC time format.
  • expire-at-timestamp: Timestamp when the key expires.

key-ttl

No

None

String

Supplementary parameter of key-ttl-mode. Available values are as follows:

  • If key-ttl-mode is no-ttl, this parameter does not need to be configured.
  • If key-ttl-mode is expire-msec, set this parameter to a string that can be parsed into the Long type. For example, 5000 indicates that the key will expire in 5000 ms.
  • If key-ttl-mode is expire-at-date, set this parameter to a date. For example, 2011-12-03T10:15:30 indicates that the expiration time is 2011-12-03 18:15:30 (UTC+8).
  • If key-ttl-mode is expire-at-timestamp, set this parameter to a timestamp, in milliseconds. For example, 1679385600000 indicates that the expiration time is 2023-03-21 16:00:00.

Example

In this example, data is read from the Kafka data source and written to the Redis result table. The procedure is as follows:

  1. Create an enhanced datasource connection in the VPC and subnet where Redis locates, and bind the connection to the required Flink elastic resource pool. For details, see Enhanced Datasource Connections.
  2. Set Redis security groups and add inbound rules to allow access from the Flink queue. Test the connectivity using the Redis address by referring to Testing Address Connectivity. If the connection is successful, the datasource is bound to the queue. Otherwise, the binding fails.
  3. Create a Flink OpenSource SQL job. Enter the following job script and submit the job.
    When you create a job, set Flink Version to 1.12 on the Running Parameters tab. Select Save Job Log, and specify the OBS bucket for saving job logs. Change the values of the parameters in bold as needed in the following script.
    CREATE TABLE orders (
      order_id string,
      order_channel string,
      order_time string,
      pay_amount double,
      real_pay double,
      pay_time string,
      user_id string,
      user_name string,
      area_id string
    ) WITH (
      'connector' = 'kafka',
      'topic' = '<yourTopic>',
      'properties.bootstrap.servers' = '<yourKafka>:<port>',
      'properties.group.id' = '<yourGroupId>',
      'scan.startup.mode' = 'latest-offset',
      'format' = 'json'
    );
    --In the following redisSink table, data-type is set to default value hash, schema-syntax is fields, and order_id is defined as the primary key. Therefore, the value of this field is used as the Redis key.
    CREATE TABLE redisSink (
      order_id string,
      order_channel string,
      order_time string,
      pay_amount double,
      real_pay double,
      pay_time string,
      user_id string,
      user_name string,
      area_id string,
      primary key (order_id) not enforced
    ) WITH (
      'connector' = 'redis',
      'host' = '<yourRedis>',
      'password' = '<yourPassword>',
      'deploy-mode' = 'master-replica',
      'schema-syntax' = 'fields'
    );
    
    insert into redisSink select * from orders;
  4. Connect to the Kafka cluster and insert the following test data into Kafka:
    {"order_id":"202103241000000001", "order_channel":"webShop", "order_time":"2021-03-24 10:00:00", "pay_amount":"100.00", "real_pay":"100.00", "pay_time":"2021-03-24 10:02:03", "user_id":"0001", "user_name":"Alice", "area_id":"330106"}
    
    {"order_id":"202103241606060001", "order_channel":"appShop", "order_time":"2021-03-24 16:06:06", "pay_amount":"200.00", "real_pay":"180.00", "pay_time":"2021-03-24 16:10:06", "user_id":"0001", "user_name":"Alice", "area_id":"330106"}
  5. Run the following commands in Redis and view the result:
    • Obtain the result whose key is 202103241606060001.

      Run following command:

      HGETALL 202103241606060001
      Command output:
       1) "user_id"
       2) "0001"
       3) "user_name"
       4) "Alice"
       5) "pay_amount"
       6) "200.0"
       7) "real_pay"
       8) "180.0"
       9) "order_time"
      10) "2021-03-24 16:06:06"
      11) "area_id"
      12) "330106"
      13) "order_channel"
      14) "appShop"
      15) "pay_time"
      16) "2021-03-24 16:10:06"
    • Obtain the result whose key is 202103241000000001.

      Run following command:

      HGETALL 202103241000000001
      Command output:
       1) "user_id"
       2) "0001"
       3) "user_name"
       4) "Alice"
       5) "pay_amount"
       6) "100.0"
       7) "real_pay"
       8) "100.0"
       9) "order_time"
      10) "2021-03-24 10:00:00"
      11) "area_id"
      12) "330106"
      13) "order_channel"
      14) "webShop"
      15) "pay_time"
      16) "2021-03-24 10:02:03"

FAQ

  • Q: When data-type is set, why is the final result data less than the input data?

    A: This is because the input data contains duplicate data. Deduplication is performed in the Redis set, and the number of records in the result decreases.

  • Q: What should I do if Flink job logs contain the following error information?
    org.apache.flink.table.api.ValidationException: SQL validation failed. From line 1, column 40 to line 1, column 105: Parameters must be of the same type

    A: The array type is used. However, the types of fields in the array are different. You need to ensure that the types of fields in the array in Redis are the same.

  • Q: What should I do if Flink job logs contain the following error information?
    org.apache.flink.addons.redis.core.exception.RedisConnectorException: Wrong Redis schema for 'map' syntax: There should be a key (possibly) and 1 MAP non-key column.

    A: When schema-syntax is map, the table creation statement in Flink can contain only one non-primary key column, and the column type must be map.

  • Q: What should I do if Flink job logs contain the following error information?
    org.apache.flink.addons.redis.core.exception.RedisConnectorException: Wrong Redis schema for 'array' syntax: There should be a key (possibly) and 1 ARRAY non-key column.

    A: When schema-syntax is array, the table creation statement in Flink can contain only one non-primary key column, and the column type must be array.

  • Q: What is the function of schema-syntax since data-type has been set?

    A: schema-syntax is used to process special types, such as map and array.

    • If it is set to fields, the value of each field is processed. If it is set to array or map, each element in the field is processed. For fields, the field value of the map or array type is directly used as a value in Redis.
    • For array or map, each value in the array is used as a Redis value, and the field value of the map is used as the Redis value. array-scores is used to process the sorted-set data type. It indicates that two array fields are used, the first one is the value in the set, and the second one is the score. fields-scores is used to process the sorted-set data type, indicating that the score is derived from the defined field. The field of an odd number except the primary key indicates the value in the set, and its next field indicates its score. Therefore, its next field must be of the double type.
  • Q: If data-type is hash, what are the differences between schema-syntax set to fields and that to map?

    A: When fields is used, the field name in Flink is used as the Redis field of the hash data type, and the value of that field is used as the value of the hash data type in Redis. When map is used, the field key in Flink is used as the Redis field of the hash data type, and the value of that field is used as the value of the hash data type in Redis. The following is an example:

    • For fields:
      1. The execution script of the Flink job is as follows:
        CREATE TABLE orders (
          order_id string,
          order_channel string,
          order_time string,
          pay_amount double,
          real_pay double,
          pay_time string,
          user_id string,
          user_name string,
          area_id string
        ) WITH (
          'connector' = 'kafka',
          'topic' = 'kafkaTopic',
          'properties.bootstrap.servers' = 'KafkaAddress1:KafkaPort,KafkaAddress2:KafkaPort',
          'properties.group.id' = 'GroupId',
          'scan.startup.mode' = 'latest-offset',
          'format' = 'json'
        );
        
        CREATE TABLE redisSink (
          order_id string,
          maptest Map<string, String>,
          primary key (order_id) not enforced
        ) WITH (
          'connector' = 'redis',
          'host' = 'RedisIP',
          'password' = 'RedisPassword',
          'deploy-mode' = 'master-replica',
          'schema-syntax' = 'fields'
        );
        
        insert into redisSink select order_id, Map[user_id, area_id] from orders;
      2. Connect to the Kafka cluster and insert the following test data into the Kafka topic:
        {"order_id":"202103241000000001", "order_channel":"webShop", "order_time":"2021-03-24 10:00:00", "pay_amount":"100.00", "real_pay":"100.00", "pay_time":"2021-03-24 10:02:03", "user_id":"0001", "user_name":"Alice", "area_id":"330106"}
      3. In the Redis, the result is as follows:
        1) "maptest"
        2) "{0001=330106}"
    • For map:
      1. The execution script of the Flink job is as follows:
        CREATE TABLE orders (
          order_id string,
          order_channel string,
          order_time string,
          pay_amount double,
          real_pay double,
          pay_time string,
          user_id string,
          user_name string,
          area_id string
        ) WITH (
          'connector' = 'kafka',
          'topic' = 'kafkaTopic',
          'properties.bootstrap.servers' = 'KafkaAddress1:KafkaPort,KafkaAddress2:KafkaPort',
          'properties.group.id' = 'GroupId',
          'scan.startup.mode' = 'latest-offset',
          'format' = 'json'
        );
        
        CREATE TABLE redisSink (
          order_id string,
          maptest Map<string, String>,
          primary key (order_id) not enforced
        ) WITH (
          'connector' = 'redis',
          'host' = 'RedisIP',
          'password' = 'RedisPassword',
          'deploy-mode' = 'master-replica',
          'schema-syntax' = 'map'
        );
        
        insert into redisSink select order_id, Map[user_id, area_id] from orders;
      2. Connect to the Kafka cluster and insert the following test data into the Kafka topic:
        {"order_id":"202103241000000001", "order_channel":"webShop", "order_time":"2021-03-24 10:00:00", "pay_amount":"100.00", "real_pay":"100.00", "pay_time":"2021-03-24 10:02:03", "user_id":"0001", "user_name":"Alice", "area_id":"330106"}
      3. In the Redis, the result is as follows:
        1) "0001"
        2) "330106"
  • Q: If data-type is list, what are the differences between schema-syntax set to fields and that to array?

    A: The setting to fields or array does not result in different results. The only difference is that in the Flink table creation statement. fields can be multiple fields. However, array requires that the field is of the array type and the data types in the array must be the same. Therefore, fields are more flexible.

    • For fields:
      1. The execution script of the Flink job is as follows:
        CREATE TABLE orders (
          order_id string,
          order_channel string,
          order_time string,
          pay_amount double,
          real_pay double,
          pay_time string,
          user_id string,
          user_name string,
          area_id string
        ) WITH (
          'connector' = 'kafka',
          'topic' = 'kafkaTopic',
          'properties.bootstrap.servers' = 'KafkaAddress1:KafkaPort,KafkaAddress2:KafkaPort',
          'properties.group.id' = 'GroupId',
          'scan.startup.mode' = 'latest-offset',
          'format' = 'json'
        );
        
        CREATE TABLE redisSink (
          order_id string,
          order_channel string,
          order_time string,
          pay_amount double,
          real_pay double,
          pay_time string,
          user_id string,
          user_name string,
          area_id string, 
          primary key (order_id) not enforced
        ) WITH (
          'connector' = 'redis',
          'host' = 'RedisIP',
          'password' = 'RedisPassword',
          'data-type' = 'list',
          'deploy-mode' = 'master-replica',
          'schema-syntax' = 'fields'
        );
        
        insert into redisSink select * from orders;
      2. Connect to the Kafka cluster and insert the following test data into the Kafka topic:
        {"order_id":"202103241000000001", "order_channel":"webShop", "order_time":"2021-03-24 10:00:00", "pay_amount":"100.00", "real_pay":"100.00", "pay_time":"2021-03-24 10:02:03", "user_id":"0001", "user_name":"Alice", "area_id":"330106"}
      3. View the result.

        Run the following command in Redis:

        LRANGE 202103241000000001 0 8
        The command output is as follows:
        1) "webShop"
        2) "2021-03-24 10:00:00"
        3) "100.0"
        4) "100.0"
        5) "2021-03-24 10:02:03"
        6) "0001"
        7) "Alice"
        8) "330106"
    • For array:
      1. The execution script of the Flink job is as follows:
        CREATE TABLE orders (
          order_id string,
          order_channel string,
          order_time string,
          pay_amount double,
          real_pay double,
          pay_time string,
          user_id string,
          user_name string,
          area_id string
        ) WITH (
          'connector' = 'kafka',
          'topic' = 'kafkaTopic',
          'properties.bootstrap.servers' = 'KafkaAddress1:KafkaPort,KafkaAddress2:KafkaPort',
          'properties.group.id' = 'GroupId',
          'scan.startup.mode' = 'latest-offset',
          'format' = 'json'
        );
        
        CREATE TABLE redisSink (
          order_id string,
          arraytest Array<String>,
          primary key (order_id) not enforced
        ) WITH (
          'connector' = 'redis',
          'host' = 'RedisIP',
          'password' = 'RedisPassword',
          'data-type' = 'list',
          'deploy-mode' = 'master-replica',
          'schema-syntax' = 'array'
        );
        
        insert into redisSink select order_id, array[order_channel,order_time,pay_time,user_id,user_name,area_id] from orders;
      2. Connect to the Kafka cluster and insert the following test data into the Kafka topic:
        {"order_id":"202103241000000001", "order_channel":"webShop", "order_time":"2021-03-24 10:00:00", "pay_amount":"100.00", "real_pay":"100.00", "pay_time":"2021-03-24 10:02:03", "user_id":"0001", "user_name":"Alice", "area_id":"330106"}
      3. In Redis, view the result. (The result is different from that of fields because data of the double type is not added to the table creation statement of the sink in Flink. Therefore, two values are missing. This is not caused by the difference between fields and array.)
        1) "webShop"
        2) "2021-03-24 10:00:00"
        3) "2021-03-24 10:02:03"
        4) "0001"
        5) "Alice"
        6) "330106"