Version: 13.x (Current)

Debezium Server CDC

This page provides instructions on how to configure the Debezium Server to connect with all the databases that are supported.

Overview

As previously introduced, in the Mia-Platform Marketplace are available a set of plugins that repackages Debezium Server image, adding the JMX Prometheus Exporter to expose collected metrics regarding the connector.

These plugins are shipped with a predefined configuration tailored for the database of interest to speed up the configuration and deployment phase. This configuration, saved in a file named application.properties is mounted as a config map on the path /debezium/conf within the container. This leverages the mechanism of Quarkus framework, on which Debezium Server is based, in order to extend the base configuration of a service.

info

Current Debezium Server plugin requires a dependency on Redis for storing the metadata regarding the offset and schema management. More details on offset management can be found below.

For more details on how Debezium works we recommend checking out the official Debezium documentation, in particular the section regarding Debezium Server.

Once the plugin is created in Mia-Platform Console, there are a few environment variables (secret or public, depending on their sensitivity) to be created and configured, which are reported in the following table.

Name	Required	Description	Default Value	Configuration Section
HTTP_PORT	false	Port exposed by the service	3000	Service Environment Variables
METRICS_HTTP_PORT	false	Port exposed by the service for metrics	3001	Service Environment Variables
LOG_LEVEL	false	Log level used by the service	INFO	Service Environment Variables
REDIS_HOSTNAME	true	Hostname or IP address to which Debezium should connect to	-	Public Variables/Project Environment Variables
REDIS_PORT	false	Redis port exposed by selected host	6379	Service Environment Variables
DEBEZIUM_SOURCE_TASK_NAME	true	Name of the specific connector task	-	Public Variables/Project Environment Variables
DEBEZIUM_REDIS_USERNAME	true	Debezium account username for accessing Redis	-	Public Variables/Project Environment Variables
DEBEZIUM_REDIS_PASSWORD	true	Debezium account password for accessing Redis	-	Public Variables/Project Environment Variables
DATABASE_HOSTNAME	true	Hostname or IP address of the database Debezium should connect and monitor changes	-	Public Variables/Project Environment Variables
DATABASE_PORT	true	Database port on which Debezium should connect to	-	Public Variables/Project Environment Variables
DEBEZIUM_DB_USERNAME	true	Debezium account username for accessing the database	-	Public Variables/Project Environment Variables
DEBEZIUM_DB_PASSWORD	true	Debezium account password for accessing the database	-	Public Variables/Project Environment Variables
TOPIC_PREFIX	true	Prefix to be added to each table on the sink topics where database changes are published	-	Public Variables/Project Environment Variables
DATABASES_LIST	true	Comma separated list of database names or regular expression that defines which databases should be monitored	-	Public Variables/Project Environment Variables
TABLES_LIST	true	Comma separated list of tables names (with their schema) or regular expression that defines which tables should be monitored	-	Public Variables/Project Environment Variables
KAFKA_BROKERS	true	Comma separated list of nodes address belonging to a Kafka cluster	-	Public Variables/Project Environment Variables
KAFKA_USERNAME	true	The Kafka username	-	Public Variables/Project Environment Variables
KAFKA_PASSWORD	true	The Kafka password	-	Public Variables/Project Environment Variables
KAFKA_CLIENT_ID	false	Client identifier employed by this application	-	Public Variables/Project Environment Variables

Debezium Server Configuration

In this section, some configuration examples for each supported database and some instructions on how to configure and enable the database itself for CDC operations are provided. Furthermore, it is clarified why Debezium Server CDC currently requires to be connected also to an instance of Redis.

In the next paragraphs, the configurations that have been tested out with the Debezium Server plugin are detailed.
Currently, the default sink configuration is prefilled with Kafka properties, since it is the main event streaming platform supported by Fast-Data. Nonetheless, it is possible to fill it with any of the Debezium-supported sink configurations as explained in the official documentation.

MySQL

CDC on MySQL can exploit the database binlog (Binary Log), which keep tracks of all the changes occurring within the database. In order to allow Debezium execute its replication operations it is therefore necessary to enable the binlog and create a user which can read the database of interest.

Regarding enabling MySQL binary log, we recommend checking out Debezium instructions and refer to the MySQL manual since it is necessary to edit the server configuration files.

Database User and Permission Configuration

The commands to create a dedicated user to CDC operations, granting the proper permissions, are provided here below.

-- CDC Configurations (replace variables identified with double curly braces with their correct value)
CREATE USER '{{DEBEZIUM_DB_USERNAME}}' IDENTIFIED WITH mysql_native_password BY '{{DEBEZIUM_DB_PASSWORD}}';

-- these are global privileges that must be granted on all the databases to be monitored
GRANT SELECT,RELOAD,SHOW DATABASES,REPLICATION SLAVE,REPLICATION CLIENT ON *.* TO '{{DEBEZIUM_DB_USERNAME}}';

-- ensure privileges are set
FLUSH PRIVILEGES;

Debezium Service Configuration

Once the database is configured, it is possible to configure the Debezium Server instance, filling in the source and sink configuration. An example using MySQL as source, Redis as offsets and metadata store and Kafka as sink is provided here below.

Click here to show/hide the example configuration for MySQL

# Application configuration
quarkus.http.port = ${HTTP_PORT}
quarkus.log.level = ${LOG_LEVEL}
# unfortunately logging in JSON returns an output with too many spacing that hinder readability
quarkus.log.console.json = false
quarkus.log.console.json.excluded-keys = sequence,loggerClassName

# Source configuration
debezium.source.name = {{DEBEZIUM_SOURCE_TASK_NAME}}
debezium.source.tasks.max = 1
debezium.source.connector.class = io.debezium.connector.mysql.MySqlConnector

# Metadata management
debezium.source.offset.flush.interval.ms = 15000
debezium.source.offset.storage = io.debezium.storage.redis.offset.RedisOffsetBackingStore
debezium.source.offset.storage.redis.address = {{REDIS_HOSTNAME}}:{{REDIS_PORT}}
debezium.source.offset.storage.redis.user = {{DEBEZIUM_REDIS_USERNAME}}
debezium.source.offset.storage.redis.password = {{DEBEZIUM_REDIS_PASSWORD}}

# it important to prevent keys overlap so that different CDC instances
# (with different configurations) do not interfere with each other
debezium.source.offset.storage.redis.key = metadata::debezium-mysql::offsets

debezium.source.schema.history.internal = io.debezium.storage.redis.history.RedisSchemaHistory
debezium.source.schema.history.internal.redis.address = {{REDIS_HOSTNAME}}:{{REDIS_PORT}}
debezium.source.schema.history.internal.redis.user = {{DEBEZIUM_REDIS_USERNAME}}
debezium.source.schema.history.internal.redis.password = {{DEBEZIUM_REDIS_PASSWORD}}

# Database connection configuration
debezium.source.database.hostname = {{DATABASE_HOSTNAME}}
debezium.source.database.port = {{DATABASE_PORT}}
debezium.source.database.user = {{DEBEZIUM_DB_USERNAME}}
debezium.source.database.password = {{DEBEZIUM_DB_PASSWORD}}
debezium.source.database.server.id = 1
debezium.source.database.ssl.mode = CONFIGURE_ME
debezium.source.topic.prefix = {{TOPIC_PREFIX}}

debezium.source.database.include.list = {{DATABASES_LIST}}
debezium.source.table.include.list = {{TABLES_LIST}}

# Disable forwarding schema changes events
debezium.source.include.schema.changes = false
debezium.source.include.schema.comments = false

debezium.source.key.converter = org.apache.kafka.connect.json.JsonConverter
debezium.source.key.converter.schemas.enable = false

debezium.source.value.converter = org.apache.kafka.connect.json.JsonConverter
debezium.source.value.converter.schemas.enable = false

debezium.source.poll.interval.ms = 500

# Sink configuration
debezium.sink.type = kafka
debezium.sink.name = {{DEBEZIUM_SINK_TASK_NAME}}
debezium.sink.kafka.producer.bootstrap.servers={{KAFKA_BROKERS}}
debezium.sink.kafka.producer.client.id = {{KAFKA_CLIENT_ID}}

debezium.sink.kafka.producer.ssl.endpoint.identification.algorithm = HTTPS
debezium.sink.kafka.producer.security.protocol = SASL_SSL
debezium.sink.kafka.producer.sasl.mechanism = SCRAM-SHA-256
debezium.sink.kafka.producer.sasl.jaas.config = org.apache.kafka.common.security.scram.ScramLoginModule required username="{{KAFKA_USERNAME}}" password="{{KAFKA_PASSWORD}}";

debezium.sink.kafka.producer.key.serializer = org.apache.kafka.common.serialization.StringSerializer
debezium.sink.kafka.producer.value.serializer = org.apache.kafka.common.serialization.StringSerializer

debezium.sink.kafka.producer.acks = all
debezium.sink.kafka.producer.max.in.flight.requests.per.connection = 5
debezium.sink.kafka.producer.enable.idempotence = true
debezium.sink.kafka.producer.min.insync.replicas = 2
debezium.sink.kafka.producer.retries = 2147483647
debezium.sink.kafka.producer.compression.type = snappy

# Format configuration
debezium.format.key = json
debezium.format.value = json

# Transforms
debezium.transforms=Reroute
debezium.transforms.Reroute.type=io.debezium.transforms.ByLogicalTableRouter
# to all tables that are mapped onto topics
debezium.transforms.Reroute.topic.regex=(.+)
# add the suffix .ingestion
debezium.transforms.Reroute.topic.replacement=$1.ingestion
# this disable the additional parameter that is added to the key, since the service it still sending
# each table to their own topic and therefore the key is still unique within that topic
debezium.transforms.Reroute.key.enforce.uniqueness=false

note

For an in depth explanation of the particular database configuration, we suggest reading the official documentation.

Oracle DB

Debezium CDC over Oracle Database exploits the LogMiner Oracle component to read and analyze the redo logs (either the archive or the online logs). In order to allow Debezium carrying out its task the ARCHIVELOG feature has to be activated.

The instructions and references to instruct an AWS RDS for Oracle instance to be ready for CDC operations are reported here below.

Enable `ARCHIVELOG`

As a first step let's check whether the feature is already available on your database instance:

SELECT LOG_MODE FROM V$DATABASE;

LOG_MODE
------------
ARCHIVELOG

In case LOG_MODE is not set to ARCHIVELOG it means the feature is not enabled. Consequently, it has to be activated executing the following AWS RDS administration procedures:

# define log files retention
exec rdsadmin.rdsadmin_util.set_configuration('archivelog retention hours', 24);

# enable further logs to be written for supporting LogMiner operations
exec rdsadmin.rdsadmin_util.alter_supplemental_logging('ADD');

caution

In order to actually activate ARCHIVELOG feature on AWS RDS for Oracle, the database backup must be enabled (it may require a database restart). This instructs the system to retain also the redo file for the configured amount of time, so that the CDC features work as expected.

Current, one day backup retention does not introduce any new cost on the AWS plan, as explained on AWS blog.

note

For further details, please check out Debezium for Oracle documentation.

Database User and Permission Configuration

In this paragraph, the commands that create a database user dedicated to Debezium Server and that grant the appropriate permissions are listed. Please, remember to replace {{DEBEZIUM_DB_USERNAME}} and {{DEBEZIUM_DB_PASSWORD}} with the real values of your use case.

# these works to assign permissions and create user dedicated to CDC
CREATE TABLESPACE logminer_tbs DATAFILE SIZE 50M AUTOEXTEND ON MAXSIZE 10G;

CREATE USER {{DEBEZIUM_DB_USERNAME}} IDENTIFIED BY {{DEBEZIUM_DB_PASSWORD}} DEFAULT TABLESPACE logminer_tbs;
GRANT CREATE SESSION TO {{DEBEZIUM_DB_USERNAME}};
GRANT SELECT ON V$DATABASE to {{DEBEZIUM_DB_USERNAME}};

GRANT SELECT_CATALOG_ROLE TO {{DEBEZIUM_DB_USERNAME}};
GRANT EXECUTE_CATALOG_ROLE TO {{DEBEZIUM_DB_USERNAME}};

# fine grain permission assignments
GRANT SELECT ON "<schema-name>"."<table-name>" TO {{DEBEZIUM_DB_USERNAME}};

# enable to keep track of changes across all the tables columns
ALTER TABLE "<schema-name>"."<table-name>" ADD SUPPLEMENTAL LOG DATA (ALL) COLUMNS;

# permission to allow performing snapshot of the selected tables 
GRANT FLASHBACK ON "<schema-name>"."<table-name>" TO {{DEBEZIUM_DB_USERNAME}};

GRANT SELECT ANY TRANSACTION TO {{DEBEZIUM_DB_USERNAME}};
GRANT LOGMINING TO {{DEBEZIUM_DB_USERNAME}};

GRANT CREATE TABLE TO {{DEBEZIUM_DB_USERNAME}}; 
GRANT LOCK ANY TABLE TO {{DEBEZIUM_DB_USERNAME}}; 
GRANT CREATE SEQUENCE TO {{DEBEZIUM_DB_USERNAME}}; 

GRANT SELECT ON V$LOG TO {{DEBEZIUM_DB_USERNAME}};
GRANT SELECT ON V$LOGFILE TO {{DEBEZIUM_DB_USERNAME}};

GRANT SELECT ON V$ARCHIVED_LOG TO {{DEBEZIUM_DB_USERNAME}};
GRANT SELECT ON V$TRANSACTION TO {{DEBEZIUM_DB_USERNAME}};

# provide permissions to read a specific operations log
# Debezium can read from each of them, depending on its configuration (by default it reads from Archive Log)
GRANT READ ON DIRECTORY ONLINELOG_DIR TO {{DEBEZIUM_DB_USERNAME}};
GRANT READ ON DIRECTORY ARCHIVELOG_DIR TO {{DEBEZIUM_DB_USERNAME}};

## The database user MUST be set UPPERCASE otherwise the procedure won't work (it would return error invalid schema)
exec rdsadmin.rdsadmin_util.grant_sys_object('V_$LOG_HISTORY', '{{DEBEZIUM_DB_USERNAME}}', 'SELECT', true);
exec rdsadmin.rdsadmin_util.grant_sys_object('V_$LOGMNR_LOGS', '{{DEBEZIUM_DB_USERNAME}}', 'SELECT', true);
exec rdsadmin.rdsadmin_util.grant_sys_object('V_$LOGMNR_CONTENTS', '{{DEBEZIUM_DB_USERNAME}}', 'SELECT', true);
exec rdsadmin.rdsadmin_util.grant_sys_object('V_$LOGMNR_PARAMETERS', '{{DEBEZIUM_DB_USERNAME}}', 'SELECT', true);
exec rdsadmin.rdsadmin_util.grant_sys_object('V_$ARCHIVE_DEST_STATUS', '{{DEBEZIUM_DB_USERNAME}}', 'SELECT', true);

exec rdsadmin.rdsadmin_util.grant_sys_object('DBMS_LOGMNR', '{{DEBEZIUM_DB_USERNAME}}', 'EXECUTE', true);
exec rdsadmin.rdsadmin_util.grant_sys_object('DBMS_LOGMNR_D', '{{DEBEZIUM_DB_USERNAME}}', 'EXECUTE', true);

Redo Log Configuration

By default, four different log file are available on AWS RDS for Oracle, each of them set to a size of 128MB. Their size might not be sufficient for CDC operations, such that Debezium Server might issue the following warning log:

Redo logs may be sized too small using the default mining strategy, consider increasing redo log sizes to a minimum of 500MB.

Following this recommendation, the size of Redo log files should be increased, for example to 512MB each. The steps to achieve this are:

create a number of new log files, corresponding to the one initially available, each of them with the new required size
rotate the active log file until one of the newer files is selected
deactivate the previous log files so that they can be deleted without issues
drop the default log files

These operations are clearly described on the AWS RDS Documentation, so that we advise to read such guide to perform these actions on the database.

info

It is possible to find a reference for each of AWS RDS Administration tasks described above in the AWS RDS Administration documentation.

Debezium Service Configuration

Once the database is configured, it is possible to configure the Debezium Server instance, filling in the source and sink configuration. An example using Oracle as source, Redis as offsets and metadata store and Kafka as sink is provided here below.

Click here to show/hide the example configuration for Oracle

# Application Configuration
quarkus.http.port = ${HTTP_PORT}
quarkus.log.level = ${LOG_LEVEL}
# unfortunately logging in JSON returns an output with too many spacing that hinder readability
quarkus.log.console.json = false
quarkus.log.console.json.excluded-keys = sequence,loggerClassName

# Source Configuration
debezium.source.name = {{DEBEZIUM_SOURCE_TASK_NAME}}
debezium.source.tasks.max = 1
debezium.source.connector.class = io.debezium.connector.oracle.OracleConnector

# Metadata management
debezium.source.offset.flush.interval.ms = 15000
debezium.source.offset.storage = io.debezium.storage.redis.offset.RedisOffsetBackingStore
debezium.source.offset.storage.redis.address = {{REDIS_HOSTNAME}}:{{REDIS_PORT}}
debezium.source.offset.storage.redis.user = {{DEBEZIUM_REDIS_USERNAME}}
debezium.source.offset.storage.redis.password = {{DEBEZIUM_REDIS_PASSWORD}}

# it important to prevent keys overlap so that different CDC instances
# (with different configurations) do not interfere with each other 
debezium.source.offset.storage.redis.key = metadata::debezium-oracle::offsets

debezium.source.schema.history.internal = io.debezium.storage.redis.history.RedisSchemaHistory
debezium.source.schema.history.internal.redis.address = {{REDIS_HOSTNAME}}:{{REDIS_PORT}}
debezium.source.schema.history.internal.redis.user = {{DEBEZIUM_REDIS_USERNAME}}
debezium.source.schema.history.internal.redis.password = {{DEBEZIUM_REDIS_PASSWORD}}

# Database connection configuration
debezium.source.database.hostname = {{DATABASE_HOSTNAME}}
debezium.source.database.port = {{DATABASE_PORT}}
debezium.source.database.user = {{DEBEZIUM_DB_USERNAME}}
debezium.source.database.password = {{DEBEZIUM_DB_PASSWORD}}
debezium.source.database.dbname = {{DATABASE_NAME}}

debezium.source.database.server.id = 1
debezium.source.database.ssl.mode = CONFIGURE_ME

debezium.source.topic.prefix = {{TOPIC_PREFIX}}
debezium.source.topic.delimiter = .

debezium.source.schema.include.list = {{DATABASES_LIST}}
debezium.source.table.include.list = {{TABLES_LIST}}

# Disable forwarding schema changes events
debezium.source.include.schema.changes = false
debezium.source.include.schema.comments = false

debezium.source.key.converter = org.apache.kafka.connect.json.JsonConverter
debezium.source.key.converter.schemas.enable = false

debezium.source.value.converter = org.apache.kafka.connect.json.JsonConverter
debezium.source.value.converter.schemas.enable = false

debezium.source.poll.interval.ms = 500
debezium.source.max.batch.size = 2048
debezium.source.max.queue.size = 8192

debezium.source.tombstones.on.delete = true

debezium.source.log.mining.strategy = online_catalog

# Sink configuration
debezium.sink.type = kafka
debezium.sink.name = {{DEBEZIUM_SINK_TASK_NAME}}
debezium.sink.kafka.producer.bootstrap.servers={{KAFKA_BROKERS}}
debezium.sink.kafka.producer.client.id = {{KAFKA_CLIENT_ID}}

debezium.sink.kafka.producer.ssl.endpoint.identification.algorithm = HTTPS
debezium.sink.kafka.producer.security.protocol = SASL_SSL
debezium.sink.kafka.producer.sasl.mechanism = SCRAM-SHA-256
debezium.sink.kafka.producer.sasl.jaas.config = org.apache.kafka.common.security.scram.ScramLoginModule required username="{{KAFKA_USERNAME}}" password="{{KAFKA_PASSWORD}}";

debezium.sink.kafka.producer.key.serializer = org.apache.kafka.common.serialization.StringSerializer
debezium.sink.kafka.producer.value.serializer = org.apache.kafka.common.serialization.StringSerializer

debezium.sink.kafka.producer.acks = all
debezium.sink.kafka.producer.max.in.flight.requests.per.connection = 5
debezium.sink.kafka.producer.enable.idempotence = true
debezium.sink.kafka.producer.min.insync.replicas = 2
debezium.sink.kafka.producer.retries = 2147483647
debezium.sink.kafka.producer.compression.type = snappy

# Format configuration
debezium.format.key = json
debezium.format.value = json

# Transforms
debezium.transforms=Reroute
debezium.transforms.Reroute.type=io.debezium.transforms.ByLogicalTableRouter
# to all tables that are mapped onto topics
debezium.transforms.Reroute.topic.regex=(.+)
# add the suffix .ingestion
debezium.transforms.Reroute.topic.replacement=$1.ingestion
# this disable the additional parameter that is added to the key, since the service it still sending
# each table to their own topic and therefore the key is still unique within that topic
debezium.transforms.Reroute.key.enforce.uniqueness=false

note

For an in depth explanation of the particular database configuration, we suggest reading the official documentation.

PostgreSQL

On PostgreSQL in order to receive the executed operations from the DB (following the CDC pattern) you have mainly two options:

Use the decoderbufs plugin
Use the pgoutput plugin

The decoderbufs plugin is developed and mantained by the Debezium community and needs to be installed on your current PostgreSQL installation. The pgoutput plugin is a plugin developed and mantained by the PostgreSQL community and comes already installed with PostgresSQL 10+. To know more about the plugins and how to install them check out the Debezium official docs about it.

Database User and Permission Configuration

The Debezium PostgreSQL plugin needs a user with the permissions of LOGIN, REPLICATION and SELECT in all the observed tables in the schema. Here's a snippet on how to create such user:

CREATE ROLE your_debezium_user LOGIN REPLICATION PASSWORD 'your_debezium_password';

-- Usually the schema's name (your_db_schema) is `public`
GRANT SELECT ON ALL TABLES IN your_db_schema TO your_debezium_user;
ALTER DEFAULT PRIVILEGES IN SCHEMA your_db_schema GRANT SELECT ON TABLES TO your_debezium_user;

Debezium service configuration

Once the DB is ready to operate with the Debezium connector you can configure the connector to your needs. As mentioned before, the plugin is configured through the file application.properties but we went ahead and linked the necessary variables to your environment variables (please refer to the table in the overview section).

danger

Since the connection to a PostgresSQL is always with a specific database instead of a host, more than one database cannot be observed at the same time like in MySQL or Oracle. Because of that we've replaced the DATABASES_LIST env var for the DATABASE_NAME (name of the DB you want to observe) and the SCHEMAS_LIST (comma separated list of the schemas you want to observe).

If the configuration we've mapped is not enough to cover your needs you can directly modify the application.properties config map file in the connector service following the official indications of the Debezium Plugin. Mind that the plugin is based on the Debezium Server image so the declared properties must follow the convention.

Here's an example of the full configuration for the plugin:

Click here to show/hide the example configuration for PostgreSQL

# Application Configuration
quarkus.http.port = ${HTTP_PORT:3000}
quarkus.log.level = ${LOG_LEVEL:INFO}
quarkus.log.console.json = false
quarkus.log.console.json.excluded-keys = sequence,loggerClassName

# Source Configuration
debezium.source.name = {{DEBEZIUM_SOURCE_TASK_NAME}}
debezium.source.tasks.max = 1
debezium.source.connector.class = io.debezium.connector.postgresql.PostgresConnector
debezium.source.offset.flush.interval.ms = 15000
debezium.source.offset.storage = io.debezium.storage.redis.offset.RedisOffsetBackingStore
debezium.source.offset.storage.redis.address = {{REDIS_HOSTNAME}}:${REDIS_PORT}
debezium.source.offset.storage.redis.user = {{DEBEZIUM_REDIS_USERNAME}}
debezium.source.offset.storage.redis.password = {{DEBEZIUM_REDIS_PASSWORD}}
debezium.source.offset.storage.redis.key = metadata::debezium-postgres::offsets

debezium.source.schema.history.internal = io.debezium.storage.redis.history.RedisSchemaHistory
debezium.source.schema.history.internal.redis.address = {{REDIS_HOSTNAME}}:${REDIS_PORT}
debezium.source.schema.history.internal.redis.user = {{DEBEZIUM_REDIS_USERNAME}}
debezium.source.schema.history.internal.redis.password = {{DEBEZIUM_REDIS_PASSWORD}} 

debezium.source.database.hostname = {{DATABASE_HOSTNAME}}
debezium.source.database.port = {{DATABASE_PORT}}
debezium.source.database.user = {{DEBEZIUM_DB_USERNAME}}
debezium.source.database.password = {{DEBEZIUM_DB_PASSWORD}}

debezium.source.database.server.id = 1
debezium.source.database.ssl.mode = <CONFIGURE_ME>
debezium.source.topic.prefix = {{TOPIC_PREFIX}}
debezium.source.topic.naming.strategy = io.debezium.schema.DefaultTopicNamingStrategy

debezium.source.database.dbname = {{DATABASE_NAME}}
debezium.source.table.include.list = {{TABLES_LIST}}

debezium.source.include.schema.changes = false
debezium.source.include.schema.comments = false

debezium.source.key.converter = org.apache.kafka.connect.json.JsonConverter
debezium.source.key.converter.schemas.enable = false

debezium.source.value.converter = org.apache.kafka.connect.json.JsonConverter
debezium.source.value.converter.schemas.enable = false

debezium.source.poll.interval.ms = 500
debezium.source.max.batch.size = 2048
debezium.source.max.queue.size = 8192

debezium.source.tombstones.on.delete = true

# Sink Configuration
debezium.sink.type = kafka
debezium.sink.name = {{DEBEZIUM_SINK_TASK_NAME}}
debezium.sink.kafka.producer.bootstrap.servers={{KAFKA_BROKERS}}
debezium.sink.kafka.producer.client.id = dz-cdc-mysql-producer-{{ENVIRONMENT_TO_DEPLOY}}

debezium.sink.kafka.producer.ssl.endpoint.identification.algorithm = HTTPS
debezium.sink.kafka.producer.security.protocol = SASL_SSL
debezium.sink.kafka.producer.sasl.mechanism = SCRAM-SHA-256
debezium.sink.kafka.producer.sasl.jaas.config = org.apache.kafka.common.security.scram.ScramLoginModule required username=\"{{KAFKA_USERNAME}}\" password=\"{{KAFKA_PASSWORD}}\";

debezium.sink.kafka.producer.key.serializer = org.apache.kafka.common.serialization.StringSerializer
debezium.sink.kafka.producer.value.serializer = org.apache.kafka.common.serialization.StringSerializer

debezium.sink.kafka.producer.acks = all
debezium.sink.kafka.producer.max.in.flight.requests.per.connection = 5
debezium.sink.kafka.producer.enable.idempotence = true
debezium.sink.kafka.producer.min.insync.replicas = 2
debezium.sink.kafka.producer.retries = 2147483647
debezium.sink.kafka.producer.compression.type = snappy

# Format Configuration
debezium.format.key = json
debezium.format.value = json

# Transforms
debezium.transforms=Reroute
debezium.transforms.Reroute.type=io.debezium.transforms.ByLogicalTableRouter
debezium.transforms.Reroute.topic.regex=(.+)
debezium.transforms.Reroute.topic.replacement=$1.ingestion
debezium.transforms.Reroute.key.enforce.uniqueness=false

DB2

In order to receive the executed operations from the DB2 database (following the CDC pattern) you have to enable Change Data Capture on your database. To do that, you have to satisfy these requirements:

You are logged in to DB2 as the db2instl user.
On the DB2 host, the Debezium management UDFs are available in the $HOME/asncdctools/src directory. UDFs are available from the Debezium examples repository.

info

The DB2 connector has been tested with a DB2 Docker container, so we will report info based on that. If you would like to get deeper insights on how the database works, please check the official documentation.

note

If you have a docker instance of DB2, you can easily access as superuser with

docker exec -ti db2server bash -c "su - db2inst1"

We also suggest creating a DB2 sample with db2sampl command as it has been done here to actually create some files that could be needed for later steps. This is just for test purposes and the sample database could (and should) be dropped after the test is concluded successfully.

Database User and Permission Configuration

Supposing you already have a DB2 instance running with the requirements explained above and superuser privileges, to put your tables into Capture Mode you have to follow the following procedure:

Connect to the database DB_NAME you want to enable CDC on
```
db2 connect to $DB_NAME
```
Compile the Debezium management UDFs on the DB2 server host by using the bldrtn command provided with Db2:
```
cd $HOME/asncdctools/src
/opt/ibm/db2/V11.5/samples/c/bldrtn asncdc
```
Start the database DB_NAME you want to monitor
```
db2 start db $DB_NAME
```

Make it possible for the JDBC driver to read the DB2 metadata catalog

cd $HOME/sqllib/bnd
db2 bind db2schema.bnd blocking all grant public sqlerror continue

Make a backup of $DB_NAME, because the ASN agent must have a starting point to read from and then reconnect to it
```
db2 backup db $DB_NAME to $BACK_UP_LOCATION
db2 restart db $DB_NAME
db2 connect to $DB_NAME
```
note
You can always set BACK_UP_LOCATION=/dev/null, but it is not recommended for production environment.

Copy Debezium management UDFs and set the right permissions

cp $HOME/asncdctools/src/asncdc $HOME/sqllib/function
chmod 777 -R $HOME/sqllib/function

Enable the Debezium UDF that starts and stops the ASN capture agent
```
db2 -tvmf $HOME/asncdctools/src/asncdc_UDF.sql
```

Create the ASN control tables:

db2 -tvmf $HOME/asncdctools/src/asncdctables.sql

danger

If the command returns error this could be due to the presence of tables with the same name inside the ASNCDC schema. We suggest to verify that the schema of tables to be created corresponds to the one the script should create.

Enable the Debezium UDF that adds tables to capture mode and removes tables from capture mode:
```
db2 -tvmf $HOME/asncdctools/src/asncdcaddremove.sql
```

You can then start the ASN agent with

db2 "VALUES ASNCDC.ASNCDCSERVICES('start','asncdc');"

danger

It is very likely that you will receive this output:

/database/config/db2inst1/sqllib/bin/asncap capture_schema=asncdc capture_server=$DB_NAME &

In this case run this command before proceeding and check again if the ASN agent is up rerunning. Please note that this command will start a background process which will trigger the ASN agent.

db2 "VALUES ASNCDC.ASNCDCSERVICES('start','asncdc');"

until the output is

asncap is already running

If you receive this output at first try you can skip this part.

Actually put $SCHEMA.$TABLE into capture mode using the above defined UDFs
```
db2 "CALL ASNCDC.ADDTABLE('$SCHEMA', '$TABLE');"
```

Reinitialize the ASN service

db2 "VALUES ASNCDC.ASNCDCSERVICES('reinit','asncdc');"

note

For an in depth explanation of the steps, we suggest reading the official documentation

Debezium service configuration

Once the DB is ready to operate with the Debezium connector, you need to load the DB2 JDBC driver in the Debezium Server image. The template available in the marketplace simplifies how to do it. After this, you are ready to configure the connector to your needs. As mentioned before, the service provided by the template is configured through the file application.properties but we went ahead and linked the necessary variables to your environment variables (please refer to the table in the overview section).

danger

Since the connection to a DB2 is always with a specific database instead of a host, more than one database cannot be observed at the same time like in MySQL or Oracle, as mentioned here. Because of that we've replaced the DATABASES_LIST env var for the DATABASE_NAME (name of the DB you want to observe).

If the configuration we've mapped is not enough to cover your needs you can directly modify the application.properties config map file in the connector service following the official indications of the Debezium Plugin. Keep in mind that the template is based on the Debezium Server image so the declared properties must follow the convention.

Here's an example of the full configuration for the plugin:

Click here to show/hide the example configuration for PostgreSQL

# Application Configuration
quarkus.http.port = ${HTTP_PORT:3000}
quarkus.log.level = ${LOG_LEVEL:INFO}
quarkus.log.console.json = false
quarkus.log.console.json.excluded-keys = sequence,loggerClassName

# Source Configuration
debezium.source.name = {{DEBEZIUM_SOURCE_TASK_NAME}}
debezium.source.tasks.max = 1
debezium.source.connector.class = io.debezium.connector.db2.Db2Connector
debezium.source.offset.flush.interval.ms = 15000
debezium.source.offset.storage = io.debezium.storage.redis.offset.RedisOffsetBackingStore
debezium.source.offset.storage.redis.address = {{REDIS_HOSTNAME}}:${REDIS_PORT}
debezium.source.offset.storage.redis.user = {{DEBEZIUM_REDIS_USERNAME}}
debezium.source.offset.storage.redis.password = {{DEBEZIUM_REDIS_PASSWORD}}
debezium.source.offset.storage.redis.key = metadata::debezium-postgres::offsets

debezium.source.schema.history.internal = io.debezium.storage.redis.history.RedisSchemaHistory
debezium.source.schema.history.internal.redis.address = {{REDIS_HOSTNAME}}:${REDIS_PORT}
debezium.source.schema.history.internal.redis.user = {{DEBEZIUM_REDIS_USERNAME}}
debezium.source.schema.history.internal.redis.password = {{DEBEZIUM_REDIS_PASSWORD}} 

debezium.source.database.hostname = {{DATABASE_HOSTNAME}}
debezium.source.database.port = {{DATABASE_PORT}}
debezium.source.database.user = {{DEBEZIUM_DB_USERNAME}}
debezium.source.database.password = {{DEBEZIUM_DB_PASSWORD}}

debezium.source.database.server.id = 1
debezium.source.database.ssl.mode = <CONFIGURE_ME>
debezium.source.topic.prefix = {{TOPIC_PREFIX}}
debezium.source.topic.naming.strategy = io.debezium.schema.DefaultTopicNamingStrategy

debezium.source.database.dbname = {{DATABASE_NAME}}
debezium.source.table.include.list = {{TABLES_LIST}}

debezium.source.include.schema.changes = false
debezium.source.include.schema.comments = false

debezium.source.key.converter = org.apache.kafka.connect.json.JsonConverter
debezium.source.key.converter.schemas.enable = false

debezium.source.value.converter = org.apache.kafka.connect.json.JsonConverter
debezium.source.value.converter.schemas.enable = false

debezium.source.poll.interval.ms = 500
debezium.source.max.batch.size = 2048
debezium.source.max.queue.size = 8192

debezium.source.tombstones.on.delete = true

# Sink Configuration
debezium.sink.type = kafka
debezium.sink.name = {{DEBEZIUM_SINK_TASK_NAME}}
debezium.sink.kafka.producer.bootstrap.servers={{KAFKA_BROKERS}}
debezium.sink.kafka.producer.client.id = dz-cdc-mysql-producer-{{ENVIRONMENT_TO_DEPLOY}}

debezium.sink.kafka.producer.ssl.endpoint.identification.algorithm = HTTPS
debezium.sink.kafka.producer.security.protocol = SASL_SSL
debezium.sink.kafka.producer.sasl.mechanism = SCRAM-SHA-256
debezium.sink.kafka.producer.sasl.jaas.config = org.apache.kafka.common.security.scram.ScramLoginModule required username=\"{{KAFKA_USERNAME}}\" password=\"{{KAFKA_PASSWORD}}\";

debezium.sink.kafka.producer.key.serializer = org.apache.kafka.common.serialization.StringSerializer
debezium.sink.kafka.producer.value.serializer = org.apache.kafka.common.serialization.StringSerializer

debezium.sink.kafka.producer.acks = all
debezium.sink.kafka.producer.max.in.flight.requests.per.connection = 5
debezium.sink.kafka.producer.enable.idempotence = true
debezium.sink.kafka.producer.min.insync.replicas = 2
debezium.sink.kafka.producer.retries = 2147483647
debezium.sink.kafka.producer.compression.type = snappy

# Format Configuration
debezium.format.key = json
debezium.format.value = json

# Transforms
debezium.transforms=Reroute
debezium.transforms.Reroute.type=io.debezium.transforms.ByLogicalTableRouter
debezium.transforms.Reroute.topic.regex=(.+)
debezium.transforms.Reroute.topic.replacement=$1.ingestion
debezium.transforms.Reroute.key.enforce.uniqueness=false

note

For an in depth explanation of the particular database configuration, we suggest reading the official documentation.

Debezium Server Generic Template

If none of the databases mentioned in the previous sections are being used, it is still possible to connect to the ones supported by debezium that we don't provide as a service in the marketplace with the Debezium Server.

By referring to the official documentation, you can configure any sink configurations supported by Debezium. To set up the template configuration, you must configure the application.properties file as directed in the documentation. Additionally, the JMX Prometheus Exporter configuration must also be configured.

The Debezium Server Generic template is available in Mia Marketplace under the category Add-ons - Fast Data Connectors.

Offsets Management

When Debezium CDC starts it initially creates a snapshot of the database, if configured so, and then it starts monitoring any change that occurs over the configured databases and tables. To keep track of which change events were consumed and forwarded to the sink resource, Debezium Server requires saving the offsets metadata somewhere. Currently, these methods are supported:

on file: org.apache.kafka.connect.storage.FileOffsetBackingStore
on Redis: io.debezium.storage.redis.offset.RedisOffsetBackingStore

Considering that the Debezium Server plugin offered in Mia-Platform Marketplace is tailored for being deployed on Kubernetes as a Deployment resource, it is recommended to exploit the Redis method to not lose progress across potential pod restarts.

note

The offset management mechanism in the standalone Debezium Server differs from Debezium deployed as a Kafka Connector, since in the latter offsets metadata are handled directly by Kafka Connect framework.

Redis Storage

Here is provided a Redis (v6) configuration file that provides the set of features to support Debezium offsets management for development purposes. This configuration can be employed to deploy Redis in Kubernetes alongside the Debezium Server pod.

caution

The following Redis configuration is suitable only for development purposes. In fact, all the data stored within it are not saved back onto a disk, but only retained in memory. Therefore, any service restart would invalidate any CDC progress and cause your CDC application to restart.

Consequently, we strongly recommend to rely on an external production ready Redis cluster.

## Generic config ##
port 6379
# listen on all the interfaces for incoming connections
bind 0.0.0.0

## Disable disk persistence when deploying on Kubernetes ##
save ""
appendonly no

## Security & ACLs ##
protected-mode yes

# NOTE: +@connection is necessary to be placed after -@dangerous, otherwise the user won't have the permissions to connect to Redis
user {{DEBEZIUM_REDIS_USERNAME}} on +@all -@dangerous +@connection allkeys allchannels >{{DEBEZIUM_REDIS_PASSWORD}}

# disable default user
user default off

Overview​

Debezium Server Configuration​

MySQL​

Database User and Permission Configuration​

Debezium Service Configuration​

Oracle DB​

Enable ARCHIVELOG​

Database User and Permission Configuration​

Redo Log Configuration​

Debezium Service Configuration​

PostgreSQL​

Database User and Permission Configuration​

Debezium service configuration​

DB2​

Database User and Permission Configuration​

Debezium service configuration​

Debezium Server Generic Template​

Offsets Management​

Redis Storage​

Overview

Debezium Server Configuration

MySQL

Database User and Permission Configuration

Debezium Service Configuration

Oracle DB

Enable `ARCHIVELOG`

Database User and Permission Configuration

Redo Log Configuration

Debezium Service Configuration

PostgreSQL

Database User and Permission Configuration

Debezium service configuration

DB2

Database User and Permission Configuration

Debezium service configuration

Debezium Server Generic Template

Offsets Management

Redis Storage