Updated on 2023-11-17 GMT+08:00

HBase Result Table

Function

DLI outputs the job data to HBase. HBase is a column-oriented distributed cloud storage system that features enhanced reliability, excellent performance, and elastic scalability. It applies to the storage of massive amounts of data and distributed computing. You can use HBase to build a storage system capable of storing TB- or even PB-level data. With HBase, you can filter and analyze data with ease and get responses in milliseconds, rapidly mining data value. Structured and semi-structured key-value data can be stored, including messages, reports, recommendation data, risk control data, logs, and orders. With DLI, you can write massive volumes of data to HBase at a high speed and with low latency.

Prerequisites

  • An enhanced datasource connection has been created for DLI to connect to HBase, so that jobs can run on the dedicated queue of DLI and you can set the security group rules as required.
  • If MRS HBase is used, IP addresses of all hosts in the MRS cluster have been added to host information of the enhanced datasource connection.

    For details, see Modifying Host Information in the Data Lake Insight User Guide.

  • 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.
  • The column families in created HBase result table must be declared as the ROW type, the field names map the column family names, and the nested field names map the column qualifier names. There is no need to declare all the families and qualifiers in the schema. Users can declare what is used in the query. Except the ROW type fields, the single atomic type field (for example, STRING or BIGINT) will be recognized as the HBase rowkey. The rowkey field can be an arbitrary name, but should be quoted using backticks if it is a reserved keyword.

Syntax

create table hbaseSink (
  attr_name attr_type 
  (',' attr_name attr_type)* 
  ','PRIMARY KEY (attr_name, ...) NOT ENFORCED)
) with (
  'connector' = 'hbase-2.2',
  'table-name' = '',
  'zookeeper.quorum' = ''
);

Parameters

Table 1 Parameter description

Parameter

Mandatory

Default Value

Data Type

Description

connector

Yes

None

String

Connector to be used. Set this parameter to hbase-2.2.

table-name

Yes

None

String

Name of the HBase table to connect.

zookeeper.quorum

Yes

None

String

HBase ZooKeeper instance information, in the format of ZookeeperAddress:ZookeeperPort.

The following uses an MRS HBase cluster as an example to describe how to obtain the IP address and port number of ZooKeeper used by this parameter:

  • On MRS Manager, choose Cluster and click the name of the desired cluster. Choose Services > ZooKeeper > Instance, and obtain the IP address of the ZooKeeper instance.
  • On MRS Manager, choose Cluster and click the name of the desired cluster. Choose Services > ZooKeeper > Configurations > All Configurations, search for the clientPort parameter, and obtain its value, that is, the ZooKeeper port number.

zookeeper.znode.parent

No

/hbase

String

Root directory in ZooKeeper. The default value is /hbase.

null-string-literal

No

null

String

Representation for null values for string fields.

The HBase sink encodes/decodes empty bytes as null values for all types except the string type.

sink.buffer-flush.max-size

No

2mb

MemorySize

Maximum size in memory of buffered rows for each write request.

This can improve performance for writing data to the HBase database, but may increase the latency.

You can set this parameter to 0 to disable it.

sink.buffer-flush.max-rows

No

1000

Integer

Maximum number of rows to buffer for each write request.

This can improve performance for writing data to the HBase database, but may increase the latency.

You can set this parameter to 0 to disable it.

sink.buffer-flush.interval

No

1s

Duration

Interval for refreshing the buffer, during which data is refreshed by asynchronous threads.

This can improve performance for writing data to the HBase database, but may increase the latency.

You can set this parameter to 0 to disable it.

Note: If sink.buffer-flush.max-size and sink.buffer-flush.max-rows are both set to 0 and the buffer refresh interval is configured, the buffer is asynchronously refreshed.

The format is {length value}{time unit label}, for example, 123ms, 321s. The supported time units include d, h, min, s, and ms (default unit).

sink.parallelism

No

None

Integer

Defines the parallelism of the HBase sink operator.

By default, the parallelism is determined by the framework using the same parallelism of the upstream chained operator.

krb_auth_name

No

None

String

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

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

Data Type Mapping

HBase stores all data as byte arrays. The data needs to be serialized and deserialized during read and write operations.

When serializing and de-serializing, Flink HBase connector uses utility class org.apache.hadoop.hbase.util.Bytes provided by HBase (Hadoop) to convert Flink data types to and from byte arrays.

Flink HBase connector encodes null values to empty bytes, and decode empty bytes to null values for all data types except the string type. For the string type, the null literal is determined by the null-string-literal option.

Table 2 Data type mapping

Flink SQL Type

HBase Conversion

CHAR / VARCHAR / STRING

byte[] toBytes(String s)

String toString(byte[] b)

BOOLEAN

byte[] toBytes(boolean b)

boolean toBoolean(byte[] b)

BINARY / VARBINARY

Returns byte[] as is.

DECIMAL

byte[] toBytes(BigDecimal v)

BigDecimal toBigDecimal(byte[] b)

TINYINT

new byte[] { val }

bytes[0] // returns first and only byte from bytes

SMALLINT

byte[] toBytes(short val)

short toShort(byte[] bytes)

INT

byte[] toBytes(int val)

int toInt(byte[] bytes)

BIGINT

byte[] toBytes(long val)

long toLong(byte[] bytes)

FLOAT

byte[] toBytes(float val)

float toFloat(byte[] bytes)

DOUBLE

byte[] toBytes(double val)

double toDouble(byte[] bytes)

DATE

Stores the number of days since epoch as an int value.

TIME

Stores the number of milliseconds of the day as an int value.

TIMESTAMP

Stores the milliseconds since epoch as a long value.

ARRAY

Not supported

MAP / MULTISET

Not supported

ROW

Not supported

Example

In this example, data is read from the Kafka data source and written to the HBase result table. The procedure is as follows (the HBase versions used in this example are 1.3.1 and 2.2.3):

  1. Create an enhanced datasource connection in the VPC and subnet where HBase and Kafka locate, and bind the connection to the required Flink elastic resource pool. For details, see Enhanced Datasource Connections. Add MRS host information for the enhanced datasource connectionModifying Host Information.
  2. Set HBase and Kafka security groups and add inbound rules to allow access from the Flink queue. Test the connectivity using the HBase and Kafka address by referring to Testing Address Connectivity. If the connection is successful, the datasource is bound to the queue. Otherwise, the binding fails.
  3. Use the HBase shell to create HBase table order that has only one column family detail. For details, see Using HBase from Scratch.
    create 'order', {NAME => 'detail'}
  4. Create a Flink OpenSource SQL job. Enter the following job script and submit the job. The job script uses Kafka as the data source and HBase as the result table (the Rowkey is order_id and the column family name is detail).
    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' = 'KafkaTopic',
      'properties.bootstrap.servers' = 'KafkaAddress1:KafkaPort,KafkaAddress2:KafkaPort',
      'properties.group.id' = 'GroupId',
      'scan.startup.mode' = 'latest-offset',
      'format' = 'json'
    );
    
    create table hbaseSink(
      order_id string,
      detail Row(
        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' = 'hbase-2.2',
      'table-name' = 'order',
      'zookeeper.quorum' = 'ZookeeperAddress:ZookeeperPort',
      'sink.buffer-flush.max-rows' = '1'
    );
    
    insert into hbaseSink select order_id, Row(order_channel,order_time,pay_amount,real_pay,pay_time,user_id,user_name,area_id) from orders;
  5. Connect to the Kafka cluster and enter the following data to 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"}
    
    {"order_id":"202103251202020001", "order_channel":"miniAppShop", "order_time":"2021-03-25 12:02:02", "pay_amount":"60.00", "real_pay":"60.00", "pay_time":"2021-03-25 12:03:00", "user_id":"0002", "user_name":"Bob", "area_id":"330110"}
  6. Run the following statement on the HBase shell to view the data result:
     scan 'order'
    The data result is as follows:
    202103241000000001   column=detail:area_id, timestamp=2021-12-16T21:30:37.954, value=330106
    
    202103241000000001   column=detail:order_channel, timestamp=2021-12-16T21:30:37.954, value=webShop
    
    202103241000000001   column=detail:order_time, timestamp=2021-12-16T21:30:37.954, value=2021-03-24 10:00:00
    
    202103241000000001   column=detail:pay_amount, timestamp=2021-12-16T21:30:37.954, value=@Y\x00\x00\x00\x00\x00\x00
    
    202103241000000001   column=detail:pay_time, timestamp=2021-12-16T21:30:37.954, value=2021-03-24 10:02:03
    
    202103241000000001   column=detail:real_pay, timestamp=2021-12-16T21:30:37.954, value=@Y\x00\x00\x00\x00\x00\x00
    
    202103241000000001   column=detail:user_id, timestamp=2021-12-16T21:30:37.954, value=0001
    
    202103241000000001   column=detail:user_name, timestamp=2021-12-16T21:30:37.954, value=Alice
    
    202103241606060001   column=detail:area_id, timestamp=2021-12-16T21:30:44.842, value=330106
    
    202103241606060001   column=detail:order_channel, timestamp=2021-12-16T21:30:44.842, value=appShop
    
    202103241606060001   column=detail:order_time, timestamp=2021-12-16T21:30:44.842, value=2021-03-24 16:06:06
    
    202103241606060001   column=detail:pay_amount, timestamp=2021-12-16T21:30:44.842, value=@i\x00\x00\x00\x00\x00\x00
    
    202103241606060001   column=detail:pay_time, timestamp=2021-12-16T21:30:44.842, value=2021-03-24 16:10:06
    
    202103241606060001   column=detail:real_pay, timestamp=2021-12-16T21:30:44.842, value=@f\x80\x00\x00\x00\x00\x00
    
    202103241606060001   column=detail:user_id, timestamp=2021-12-16T21:30:44.842, value=0001
    
    202103241606060001   column=detail:user_name, timestamp=2021-12-16T21:30:44.842, value=Alice
    
    202103251202020001   column=detail:area_id, timestamp=2021-12-16T21:30:52.181, value=330110
    
    202103251202020001   column=detail:order_channel, timestamp=2021-12-16T21:30:52.181, value=miniAppShop
    
    202103251202020001   column=detail:order_time, timestamp=2021-12-16T21:30:52.181, value=2021-03-25 12:02:02
    
    202103251202020001   column=detail:pay_amount, timestamp=2021-12-16T21:30:52.181, value=@N\x00\x00\x00\x00\x00\x00
    
    202103251202020001   column=detail:pay_time, timestamp=2021-12-16T21:30:52.181, value=2021-03-25 12:03:00
    
    202103251202020001   column=detail:real_pay, timestamp=2021-12-16T21:30:52.181, value=@N\x00\x00\x00\x00\x00\x00
    
    202103251202020001   column=detail:user_id, timestamp=2021-12-16T21:30:52.181, value=0002
    
    202103251202020001   column=detail:user_name, timestamp=2021-12-16T21:30:52.181, value=Bob

FAQ

Q: What should I do if the Flink job execution fails and the log contains the following error information?

org.apache.zookeeper.ClientCnxn$SessionTimeoutException: Client session timed out, have not heard from server in 90069ms for connection id 0x0

A: The datasource connection is not bound or the binding fails. Configure the datasource connection by referring to Enhanced Datasource Connection or configure the security group of the Kafka cluster to allow access from the DLI queue.