Deploy an Enterprise HTAP Server with MariaDB Enterprise ColumnStore 5.5 and MariaDB Enterprise Server 10.5 on Ubuntu 20.04 LTS

Disable AppArmor (Debian/Ubuntu/SLES)

AppArmor must be disabled before installing MariaDB Enterprise ColumnStore.

1. Disable AppArmor:

   $ sudo systemctl disable apparmor
   

2. Reboot the system.

3. Confirm that no AppArmor profiles are loaded using aa-status:

   $ sudo aa-status
   

   Example output:

   apparmor module is loaded.
   0 profiles are loaded.
   0 profiles are in enforce mode.
   0 profiles are in complain mode.
   0 processes have profiles defined.
   0 processes are in enforce mode.
   0 processes are in complain mode.
   0 processes are unconfined but have a profile defined.
   

Create an S3 Bucket

If you want to use S3-compatible storage, it is important to create the S3 bucket before you start ColumnStore.

If you already have an S3 bucket, confirm that the bucket is empty.

We will configure Enterprise ColumnStore to use the S3 bucket later in the Configure the S3 Storage Manager section.

Install Enterprise ColumnStore via APT (Debian/Ubuntu)

1. Retrieve your Customer Download Token at https://customers.mariadb.com/downloads/token/ and substitute for customer_download_token in the following directions.

2. Configure the APT package repository.

   MariaDB Enterprise ColumnStore 5.5 is available on MariaDB Enterprise Server 10.5.

   To configure APT package repositories:

   $ sudo apt install wget
   
   $ wget https://dlm.mariadb.com/enterprise-release-helpers/mariadb_es_repo_setup
   
   $ echo "c78db828709d94876406a0ea346f13fbc38e73996795903f40e3c21385857dd4  mariadb_es_repo_setup" \
       | sha256sum -c -
   
   $ chmod +x mariadb_es_repo_setup
   
   $ sudo ./mariadb_es_repo_setup --token="customer_download_token" --apply \
      --mariadb-server-version="10.5"
   
   $ sudo apt update
   

3. Install some additional dependencies for Enterprise ColumnStore.

   On Debian 10 and Ubuntu 20.04, install the following:

   $ sudo apt install libjemalloc2
   

4. Install MariaDB Enterprise ColumnStore and package dependencies:

   $ sudo apt install mariadb-server mariadb-backup \
      libmariadb3 mariadb-client \
      mariadb-plugin-columnstore
   

5. Configure MariaDB Enterprise ColumnStore.

   Installation only loads MariaDB Enterprise ColumnStore to the system. MariaDB Enterprise ColumnStore requires configuration and additional post-installation steps before the database server is ready for use.

Enterprise Server Configuration

MariaDB Enterprise Server can be configured in the following ways:

• System variables and options can be set in a configuration file (such as /etc/my.cnf). MariaDB Enterprise Server must be restarted to apply changes made to the configuration file.

• System variables and options can be set on the command-line.

• If a system variable supports dynamic changes, then it can be set on-the-fly using the SET statement.

Configuration Files

MariaDB's packages include several bundled configuration files. It is also possible to create custom configuration files.

On Debian and Ubuntu, MariaDB's packages bundle the following configuration files:

• /etc/mysql/my.cnf

• /etc/mysql/mariadb.cnf

• /etc/mysql/mariadb.conf.d/50-client.cnf

• /etc/mysql/mariadb.conf.d/50-mysql-clients.cnf

• /etc/mysql/mariadb.conf.d/50-mysqld_safe.cnf

• /etc/mysql/mariadb.conf.d/50-server.cnf

• /etc/mysql/mariadb.conf.d/60-galera.cnf

• /etc/mysql/mariadb.conf.d/mariadb-enterprise.cnf

And on Debian and Ubuntu, custom configuration files from the following directories are read by default:

• /etc/mysql/conf.d/

• /etc/mysql/mariadb.conf.d/

Configuring MariaDB for Enterprise ColumnStore

1. Determine which system variables and options you need to configure.

   Mandatory system variables and options for HTAP include:

   System Variable/Option	Description
   character_set_server	Set this system variable to utf8
   collation_server	Set this system variable to utf8_general_ci
   columnstore_use_import_for_batchinsert	Set this system variable to ALWAYS to always use cpimport for LOAD DATA INFILE and INSERT...SELECT statements.
   server_id	Sets the numeric Server ID for this MariaDB Enterprise Server. Must be unique in the deployment.
   log_bin	Set this option to enable the Binary Log.
   binlog_format	Set this to STATEMENT for HTAP.
   log_slave_updates	Set this to OFF for HTAP.
   replicate_same_server_id	Set this to ON for HTAP.
   binlog_do_db	Set this to the name of the database that you want to replicate from InnoDB to ColumnStore.
   replicate_rewrite_db	Set this to the names of the InnoDB and ColumnStore databases in the format <innodb database>-><columnstore database>.
   replicate_wild_do_table	Set this to a pattern that matches the table names that you want to replicate from InnoDB to ColumnStore.

   Useful system variables and options for MariaDB Enterprise Server include:

   System Variable/Option	Description
   innodb_buffer_pool_size	Defines the amount of memory InnoDB reserves for the Buffer Pool. Set to no more than 16% of total memory on Servers that also run MariaDB ColumnStore.

   Useful system variables and options for HTAP include:

   System Variable/Option	Description
   relay_log	Sets the location for the Relay Log.

2. Choose a configuration file in which to configure your system variables and options.

   We recommend not making custom changes to one of the bundled configuration files. Instead, create a custom configuration file in one of the included directories. Configuration files in included directories are read in alphabetical order. If you want your custom configuration file to override the bundled configuration files, it is a good idea to prefix the custom configuration file's name with a string that will be sorted last, such as z-.

   • On Debian and Ubuntu, a good custom configuration file would be: /etc/mysql/mariadb.conf.d/z-custom-my.cnf

3. Set your system variables and options in the configuration file.

   They need to be set in a group that will be read by mariadbd, such as [mariadb] or [server].

   For example:

   [mariadb]
   log_error                              = mariadbd.err
   character_set_server                   = utf8
   collation_server                       = utf8_general_ci
   columnstore_use_import_for_batchinsert = ALWAYS
   
   # Replication Configuration (HTAP Server)
   server_id         = 1
   log_bin           = mariadb-bin
   binlog_format     = STATEMENT
   log_slave_updates = OFF
   columnstore_replication_slave    = ON
   
   # HTAP filtering rules
   
   # 1. The transactions are being
   #  replicated from itself
   replicate_same_server_id         = ON
   
   # 2. Only write queries that touch
   #   innodb_db to the binary log
   binlog_do_db            = innodb_db
   
   # 3. Rewrite innodb_db to columnstore_db
   #   prior to applying transaction
   replicate_rewrite_db    = innodb_db->columnstore_db
   
   # 4. Only replicate tables that begin with "htap"
   replicate_wild_do_table = columnstore_db.htap%
   

Configure Cross Engine Joins

When a cross engine join is executed, the ExeMgr process connects to the server using the root user with no password by default. MariaDB Enterprise Server 10.5 will reject this login attempt by default. If you plan to use Cross Engine Joins, you need to configure Enterprise ColumnStore to use a different user account and password. These directions are for configuring the cross engine join user. Directions for creating the cross engine join user are in the Create the Cross Engine Join User section.

To configure cross engine joins, perform the following steps, use the mcsSetConfig command.

For example, to configure Enterprise ColumnStore to use the cross_engine user account to connect to the server at 127.0.0.1:

$ sudo mcsSetConfig CrossEngineSupport Host 127.0.0.1
$ sudo mcsSetConfig CrossEngineSupport Port 3306
$ sudo mcsSetConfig CrossEngineSupport User cross_engine
$ sudo mcsSetConfig CrossEngineSupport Password cross_engine_passwd

Configure the S3 Storage Manager

MariaDB Enterprise ColumnStore can use S3-compatible storage, but it is not required. S3-compatible storage must be configured before it can be used.

To configure Enterprise ColumnStore to use S3-compatible storage, edit /etc/columnstore/storagemanager.cnf:

[ObjectStorage]
…
service = S3
…
[S3]
bucket = your_columnstore_bucket_name
endpoint = your_s3_endpoint
aws_access_key_id = your_s3_access_key_id
aws_secret_access_key = your_s3_secret_key
# iam_role_name = your_iam_role
# sts_region = your_sts_region
# sts_endpoint = your_sts_endpoint

[Cache]
cache_size = your_local_cache_size
path = your_local_cache_path

• The default local cache size is 2 GB.

• The default local cache path is /var/lib/columnstore/storagemanager/cache.

• Ensure that the local cache path has sufficient store space to store the local cache.

• The bucket option must be set to the name of the bucket that you created in the Create an S3 Bucket step.

• To use an IAM role, you must also uncomment and set iam_role_name, sts_region, and sts_endpoint.

Start the Enterprise ColumnStore Processes

The Enterprise Server and Enterprise ColumnStore processes can be started using the systemctl command. In case the processes were started during the installation process, use the restart command to ensure that the processes pick up the new configuration. Perform the following procedure.

1. Start the MariaDB Enterprise Server process and configure it to start automatically:

   $ sudo systemctl restart mariadb
   $ sudo systemctl enable mariadb
   

2. Start the MariaDB Enterprise ColumnStore processes and configure them to start automatically:

   $ sudo systemctl restart mariadb-columnstore
   $ sudo systemctl enable mariadb-columnstore
   

Create the Cross Engine Join User

The credentials for cross engine joins were previously configured in the Cross Engine Joins section. The user account must also be created and granted the necessary privileges to access data.

1. Connect to the server using MariaDB Client using the root@localhost user account:

   $ sudo mariadb
   

2. Create the user account with the CREATE USER statement:

   CREATE USER 'cross_engine'@'127.0.0.1'
      IDENTIFIED BY "cross_engine_passwd";
   
   CREATE USER 'cross_engine'@'localhost'
      IDENTIFIED BY "cross_engine_passwd";
   

   Note

   Please choose the same user name and password that was configured in the Cross Engine Joins section.

3. Grant the user account SELECT privileges on all databases with the GRANT statement:

   GRANT SELECT ON *.*
      TO 'cross_engine'@'127.0.0.1';
   
   GRANT SELECT ON *.*
      TO 'cross_engine'@'localhost';
   

Create the Replication User

The deployment requires MariaDB Replication, so the replication user account must be created, and the user account must be granted sufficient privileges to perform replication.

1. Connect to the server using MariaDB Client using the root@localhost user account:

   $ sudo mariadb
   

2. Create the user account with the CREATE USER statement:

   CREATE USER 'htap_replication'@'localhost'
      IDENTIFIED BY 'passwd';
   

   Note

   Please choose a password that meets your organization's password policies. If your MariaDB Enterprise Server instance has a password validation plugin installed, then the password should also meet the configured requirements.

   By default, MariaDB Enterprise Server installs the simple_password_check plugin. Its requirements are configured by the simple_password_check_digits, simple_password_check_letters_same_case, simple_password_check_minimal_length, and simple_password_check_other_characters system variables.

3. Grant the user account REPLICATION SLAVE and the BINLOG MONITOR global privileges with the GRANT statement:

   GRANT REPLICATION SLAVE, BINLOG MONITOR ON *.*
      TO 'htap_replication'@'localhost';
   

4. This user account will be used to configure replication in the Configure MariaDB Replication section.

Configure AppArmor (Debian/Ubuntu/SLES)

We disabled AppArmor in the Disable AppArmor section, but we have to create an AppArmor profile for Enterprise ColumnStore before re-enabling it. This will ensure that AppArmor does not interfere with Enterprise ColumnStore's functionality.

For information on how to create a profile, see How to create an AppArmor Profile on ubuntu.com.

Enterprise Server Administration

MariaDB Enterprise Server uses systemctl to start and stop the server processes:

Enterprise ColumnStore Administration

MariaDB Enterprise ColumnStore uses systemctl to start and stop the ColumnStore processes:

Using HTAP Replication

1. Connect to the server using MariaDB Client using the root@localhost user account:

   $ sudo mariadb
   

2. Create the databases for the InnoDB and ColumnStore tables using the CREATE DATABASE statement:

   CREATE DATABASE columnstore_db;
   
   CREATE DATABASE innodb_db;
   

3. Create the InnoDB versions of the HTAP tables using the CREATE TABLE statement:

   USE innodb_db;
   
   CREATE TABLE htap_test1 (id INT) ENGINE = InnoDB;
   
   CREATE TABLE htap_test2 (id INT) ENGINE = InnoDB;
   

4. Confirm that the tables were replicated using the SHOW TABLES statement:

   SHOW TABLES
      FROM columnstore_db;
   

   +--------------------------+
   | Tables_in_columnstore_db |
   +--------------------------+
   | htap_test1               |
   | htap_test2               |
   +--------------------------+
   

5. The replication initially creates empty InnoDB tables, which need to be transformed into ColumnStore tables and which need to be populated with the initial copy of the data:

   DROP TABLE IF EXISTS columnstore_db.htap_test1;
   
   CREATE TABLE columnstore_db.htap_test1
     ENGINE=COLUMNSTORE
     SELECT * FROM innodb_db.htap_test1;
   
   DROP TABLE IF EXISTS columnstore_db.htap_test2;
   
   CREATE TABLE columnstore_db.htap_test2
     ENGINE=COLUMNSTORE
     SELECT * FROM innodb_db.htap_test2;
   

6. Insert data into the InnoDB versions of the HTAP tables using the INSERT statement:

   USE innodb_db;
   
   INSERT INTO htap_test1
      VALUES (100);
   
   INSERT INTO htap_test2
      VALUES (200);
   

7. Confirm that the data was replicated using the SELECT statement:

   SELECT * FROM columnstore_db.htap_test1;
   

   +------+
   | id   |
   +------+
   |  100 |
   +------+
   

   SELECT * FROM columnstore_db.htap_test2;
   

   +------+
   | id   |
   +------+
   |  200 |
   +------+
   

8. Create an InnoDB table that will not be replicated:

   USE innodb_db;
   CREATE TABLE transactional_test1 (id INT) ENGINE = InnoDB;
   

9. Confirm that the table was not replicated:

   SHOW TABLES
      FROM columnstore_db
      LIKE 'transactional_%';
   

   Empty set (0.02 sec)
   

10. Create a ColumnStore table that will not be replicated:

    USE columnstore_db;
    CREATE TABLE analytical_test1 (id INT) ENGINE = ColumnStore;
    

11. Confirm that the table was not replicated:

    SHOW TABLES
       FROM innodb_db
       LIKE 'analytical_%';
    

    Empty set (0.02 sec)
    

Checking Server Status

1. Connect to the HTAP server using the MariaDB Client using the root@localhost user account:

   $ sudo mariadb
   Welcome to the MariaDB monitor.  Commands end with ; or \g.
   Your MariaDB connection id is 38
   Server version: 10.5.8-5-MariaDB-Enterprise MariaDB Enterprise Server
   
   Copyright (c) 2000, 2018, Oracle, MariaDB Corporation Ab and others.
   
   Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.
   
   MariaDB [(none)]>
   

Checking Replication Status

1. Connect to the HTAP server using the MariaDB Client using the root@localhost user account:

   $ sudo mariadb
   Welcome to the MariaDB monitor.  Commands end with ; or \g.
   Your MariaDB connection id is 38
   Server version: 10.5.8-5-MariaDB-Enterprise MariaDB Enterprise Server
   
   Copyright (c) 2000, 2018, Oracle, MariaDB Corporation Ab and others.
   
   Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.
   
   MariaDB [(none)]>
   

2. Execute the SHOW REPLICA STATUS statement.

   This statement returns a lot of columns, but the important ones are Slave_IO_Running and Slave_SQL_Running. If both of those say Yes, then replication is running correctly.

   SHOW REPLICA STATUS\G
   
   *************************** 1. row ***************************
                   Slave_IO_State: Waiting for master to send event
                      Master_Host: localhost
                      Master_User: htap_replication
                      Master_Port: 3306
                    Connect_Retry: 60
                  Master_Log_File: mariadb-bin.000002
              Read_Master_Log_Pos: 695
                   Relay_Log_File: mysqld-relay-bin.000002
                    Relay_Log_Pos: 996
            Relay_Master_Log_File: mariadb-bin.000002
                 Slave_IO_Running: Yes
                Slave_SQL_Running: Yes
                  Replicate_Do_DB:
              Replicate_Ignore_DB:
               Replicate_Do_Table:
           Replicate_Ignore_Table:
          Replicate_Wild_Do_Table: columnstore_db.htap%
      Replicate_Wild_Ignore_Table:
                       Last_Errno: 0
                       Last_Error:
                     Skip_Counter: 0
              Exec_Master_Log_Pos: 695
                  Relay_Log_Space: 1306
                  Until_Condition: None
                   Until_Log_File:
                    Until_Log_Pos: 0
               Master_SSL_Allowed: No
               Master_SSL_CA_File:
               Master_SSL_CA_Path:
                  Master_SSL_Cert:
                Master_SSL_Cipher:
                   Master_SSL_Key:
            Seconds_Behind_Master: 0
    Master_SSL_Verify_Server_Cert: No
                    Last_IO_Errno: 0
                    Last_IO_Error:
                   Last_SQL_Errno: 0
                   Last_SQL_Error:
      Replicate_Ignore_Server_Ids:
                 Master_Server_Id: 1
                   Master_SSL_Crl:
               Master_SSL_Crlpath:
                       Using_Gtid: Slave_Pos
                      Gtid_IO_Pos: 0-1-7
          Replicate_Do_Domain_Ids:
      Replicate_Ignore_Domain_Ids:
                    Parallel_Mode: optimistic
                        SQL_Delay: 0
              SQL_Remaining_Delay: NULL
          Slave_SQL_Running_State: Slave has read all relay log; waiting for more updates
                 Slave_DDL_Groups: 1
   Slave_Non_Transactional_Groups: 0
       Slave_Transactional_Groups: 1
   

Cross-Database Queries

This implementation relies on replicate_rewrite_db, so it does not support cross-database queries with statement-based replication.

For example, if the replicated database is selected by the USE, then the query will replicate properly:

USE innodb_db;

INSERT INTO htap_test1
   VALUES (100);

SELECT * FROM columnstore_db.htap_test1;

+------+
| id   |
+------+
|  100 |
+------+

However, if the replicated database is not selected, and it is instead prefixed the table name in the query, then the query will not replicate properly:

USE columnstore_db;

INSERT INTO innodb_db.htap_test1
   VALUES (200);

SELECT * FROM columnstore_db.htap_test1;

+------+
| id   |
+------+
|  100 |
+------+

Semi-Synchronous Replication

This implementation has not been tested with semi-synchronous replication.

Parallel Replication

This implementation has not been tested with parallel replication.

Row-Based Replication

This implementation requires the binlog_format system variable to be set to STATEMENT. Row-based replication is not currently supported.