Chapter 3. Database Images

The first time you use the shared volume, the database is created along with the database administrator user and the MySQL root user (if you specify the MYSQL_ROOT_PASSWORD environment variable). Afterwards, the MySQL daemon starts up. If you are re-attaching the volume to another container, then the database, database user, and the administrator user are not created, and the MySQL daemon starts.

The following command creates a new database pod with MySQL running in a container:

$ oc new-app \
    -e MYSQL_USER=<username> \
    -e MYSQL_PASSWORD=<password> \
    -e MYSQL_DATABASE=<database_name> \
    registry.access.redhat.com/openshift3/mysql-55-rhel7

OpenShift Enterprise uses Software Collections (SCLs) to install and launch MySQL. If you want to execute a MySQL command inside of a running container (for debugging), you must invoke it using bash.

To do so, first identify the name of the pod. For example, you can view the list of pods in your current project:

$ oc get pods

Then, open a remote shell session to the pod:

$ oc rsh <pod>

When you enter the container, the required SCL is automatically enabled.

You can now run the mysql command from the bash shell to start a MySQL interactive session and perform normal MySQL operations. For example, to authenticate as the database user:

bash-4.2$ mysql -u $MYSQL_USER -p$MYSQL_PASSWORD -h $HOSTNAME $MYSQL_DATABASE
Welcome to the MySQL monitor.  Commands end with ; or \g.
Your MySQL connection id is 4
Server version: 5.5.37 MySQL Community Server (GPL)
...
mysql>

When you are finished, enter quit or exit to leave the MySQL session.

The MySQL user name, password, and database name must be configured with the following environment variables:

MySQL settings can be configured with the following environment variables:

Some of the memory-related parameters have two default values. The fixed value is used when a container does not have memory limits assigned. The other value is calculated dynamically during a container’s startup based on available memory.

The MySQL image can be run with mounted volumes to enable persistent storage for the database:

Passwords are part of the image configuration, therefore the only supported method to change passwords for the database user (MYSQL_USER) and root user is by changing the environment variables MYSQL_PASSWORD and MYSQL_ROOT_PASSWORD, respectively.

You can view the current passwords by viewing the pod or deployment configuration in the web console or by listing the environment variables with the CLI:

$ oc set env pod <pod_name> --list

Whenever MYSQL_ROOT_PASSWORD is set, it enables remote access for the root user with the given password, and whenever it is unset, remote access for the root user is disabled. This does not affect the regular user MYSQL_USER, who always has remote access. This also does not affect local access by the root user, who can always log in without a password in localhost.

Changing database passwords through SQL statements or any way other than through the environment variables aforementioned causes a mismatch between the values stored in the variables and the actual passwords. Whenever a database container starts, it resets the passwords to the values stored in the environment variables.

To change these passwords, update one or both of the desired environment variables for the related deployment configuration(s) using the oc set env command. If multiple deployment configurations utilize these environment variables, for example in the case of an application created from a template, you must update the variables on each deployment configuration so that the passwords are in sync everywhere. This can be done all in the same command:

$ oc set env dc <dc_name> [<dc_name_2> ...] \
  MYSQL_PASSWORD=<new_password> \
  MYSQL_ROOT_PASSWORD=<new_root_password>

Updating the environment variables triggers the redeployment of the database server if you have a configuration change trigger. Otherwise, you must manually start a new deployment in order to apply the password changes.

To verify that new passwords are in effect, first open a remote shell session to the running MySQL pod:

$ oc rsh <pod>

From the bash shell, verify the database user’s new password:

bash-4.2$ mysql -u $MYSQL_USER -p<new_password> -h $HOSTNAME $MYSQL_DATABASE -te "SELECT * FROM (SELECT database()) db CROSS JOIN (SELECT user()) u"

If the password was changed correctly, you should see a table like this:

+------------+---------------------+
| database() | user()              |
+------------+---------------------+
| sampledb   | user0PG@172.17.42.1 |
+------------+---------------------+

To verify the root user’s new password:

bash-4.2$ mysql -u root -p<new_root_password> -h $HOSTNAME $MYSQL_DATABASE -te "SELECT * FROM (SELECT database()) db CROSS JOIN (SELECT user()) u"

If the password was changed correctly, you should see a table like this:

+------------+------------------+
| database() | user()           |
+------------+------------------+
| sampledb   | root@172.17.42.1 |
+------------+------------------+

To set up MySQL replication, a deployment configuration is defined in the example template that defines a replication controller. For MySQL master-slave replication, two deployment configurations are needed. One deployment configuration defines the MySQL master server and second the MySQL slave servers.

To tell a MySQL server to act as the master, the command field in the container’s definition in the deployment configuration must be set to run-mysqld-master. This script acts as an alternative entrypoint for the MySQL image and configures the MySQL server to run as the master in replication.

MySQL replication requires a special user that relays data between the master and slaves. The following environment variables are defined in the template for this purpose:

kind: "DeploymentConfig"
apiVersion: "v1"
metadata:
  name: "mysql-master"
spec:
  strategy:
    type: "Recreate"
  triggers:
    - type: "ConfigChange"
  replicas: 1
  selector:
    name: "mysql-master"
  template:
    metadata:
      labels:
        name: "mysql-master"
    spec:
      volumes:
        - name: "mysql-master-data"
          persistentVolumeClaim:
            claimName: "mysql-master"
      containers:
        - name: "server"
          image: "openshift/mysql-55-centos7"
          command:
            - "run-mysqld-master"
          ports:
            - containerPort: 3306
              protocol: "TCP"
          env:
            - name: "MYSQL_MASTER_USER"
              value: "${MYSQL_MASTER_USER}"
            - name: "MYSQL_MASTER_PASSWORD"
              value: "${MYSQL_MASTER_PASSWORD}"
            - name: "MYSQL_USER"
              value: "${MYSQL_USER}"
            - name: "MYSQL_PASSWORD"
              value: "${MYSQL_PASSWORD}"
            - name: "MYSQL_DATABASE"
              value: "${MYSQL_DATABASE}"
            - name: "MYSQL_ROOT_PASSWORD"
              value: "${MYSQL_ROOT_PASSWORD}"
          volumeMounts:
            - name: "mysql-master-data"
              mountPath: "/var/lib/mysql/data"
          resources: {}
          terminationMessagePath: "/dev/termination-log"
          imagePullPolicy: "IfNotPresent"
          securityContext:
            capabilities: {}
            privileged: false
      restartPolicy: "Always"
      dnsPolicy: "ClusterFirst"

Since we claimed a persistent volume in this deployment configuration to have all data persisted for the MySQL master server, you must ask your cluster administrator to create a persistent volume that you can claim the storage from.

After the deployment configuration is created and the pod with MySQL master server is started, it will create the database defined by MYSQL_DATABASE and configure the server to replicate this database to slaves.

The example provided defines only one replica of the MySQL master server. This causes OpenShift Enterprise to start only one instance of the server. Multiple instances (multi-master) is not supported and therefore you can not scale this replication controller.

To replicate the database created by the MySQL master, a deployment configuration is defined in the template. This deployment configuration creates a replication controller that launches the MySQL image with the command field set to run-mysqld-slave. This alternative entrypoints skips the initialization of the database and configures the MySQL server to connect to the mysql-master service, which is also defined in example template.

kind: "DeploymentConfig"
apiVersion: "v1"
metadata:
  name: "mysql-slave"
spec:
  strategy:
    type: "Recreate"
  triggers:
    - type: "ConfigChange"
  replicas: 1
  selector:
    name: "mysql-slave"
  template:
    metadata:
      labels:
        name: "mysql-slave"
    spec:
      containers:
        - name: "server"
          image: "openshift/mysql-55-centos7"
          command:
            - "run-mysqld-slave"
          ports:
            - containerPort: 3306
              protocol: "TCP"
          env:
            - name: "MYSQL_MASTER_USER"
              value: "${MYSQL_MASTER_USER}"
            - name: "MYSQL_MASTER_PASSWORD"
              value: "${MYSQL_MASTER_PASSWORD}"
            - name: "MYSQL_DATABASE"
              value: "${MYSQL_DATABASE}"
          resources: {}
          terminationMessagePath: "/dev/termination-log"
          imagePullPolicy: "IfNotPresent"
          securityContext:
            capabilities: {}
            privileged: false
      restartPolicy: "Always"
      dnsPolicy: "ClusterFirst"

This example deployment configuration starts the replication controller with the initial number of replicas set to 1. You can scale this replication controller in both directions, up to the resources capacity of your account.

The pods created by the MySQL slave replication controller must reach the MySQL master server in order to register for replication. The example template defines a headless service named mysql-master for this purpose. This service is not used only for replication, but the clients can also send the queries to mysql-master:3306 as the MySQL host.

To have a headless service, the portalIP parameter in the service definition is set to None. Then you can use a DNS query to get a list of the pod IP addresses that represents the current endpoints for this service.

kind: "Service"
apiVersion: "v1"
metadata:
  name: "mysql-master"
  labels:
    name: "mysql-master"
spec:
  ports:
    - protocol: "TCP"
      port: 3306
      targetPort: 3306
      nodePort: 0
  selector:
    name: "mysql-master"
  portalIP: "None"
  type: "ClusterIP"
  sessionAffinity: "None"
status:
  loadBalancer: {}

To increase the number of members in the cluster:

$ oc scale rc mysql-slave-1 --replicas=<number>

This tells the replication controller to create a new MySQL slave pod. When a new slave is created, the slave entrypoint first attempts to contact the mysql-master service and register itself to the replication set. Once that is done, the MySQL master server sends the slave the replicated database.

When scaling down, the MySQL slave is shut down and, because the slave does not have any persistent storage defined, all data on the slave is lost. The MySQL master server then discovers that the slave is not reachable anymore, and it automatically removes it from the replication.

151113  5:06:56 InnoDB: Using Linux native AIO
151113  5:06:56  InnoDB: Warning: io_setup() failed with EAGAIN. Will make 5 attempts before giving up.
InnoDB: Warning: io_setup() attempt 1 failed.
InnoDB: Warning: io_setup() attempt 2 failed.
Waiting for MySQL to start ...
InnoDB: Warning: io_setup() attempt 3 failed.
InnoDB: Warning: io_setup() attempt 4 failed.
Waiting for MySQL to start ...
InnoDB: Warning: io_setup() attempt 5 failed.
151113  5:06:59  InnoDB: Error: io_setup() failed with EAGAIN after 5 attempts.
InnoDB: You can disable Linux Native AIO by setting innodb_use_native_aio = 0 in my.cnf
151113  5:06:59 InnoDB: Fatal error: cannot initialize AIO sub-system
151113  5:06:59 [ERROR] Plugin 'InnoDB' init function returned error.
151113  5:06:59 [ERROR] Plugin 'InnoDB' registration as a STORAGE ENGINE failed.
151113  5:06:59 [ERROR] Unknown/unsupported storage engine: InnoDB
151113  5:06:59 [ERROR] Aborting

Alternatively, increase the aio-max-nr kernel resource. The following example examines the current value of aio-max-nr and doubles it.

$ sysctl fs.aio-max-nr
fs.aio-max-nr = 1048576
# sysctl -w fs.aio-max-nr=2097152

This is a per-node resolution and lasts until the next node reboot.

The first time you use the shared volume, the database is created along with the database administrator user and the PostgreSQL postgres user (if you specify the POSTGRESQL_ADMIN_PASSWORD environment variable). Afterwards, the PostgreSQL daemon starts up. If you are re-attaching the volume to another container, then the database, the database user, and the administrator user are not created, and the PostgreSQL daemon starts.

The following command creates a new database pod with PostgreSQL running in a container:

$ oc new-app \
    -e POSTGRESQL_USER=<username> \
    -e POSTGRESQL_PASSWORD=<password> \
    -e POSTGRESQL_DATABASE=<database_name> \
    registry.access.redhat.com/rhscl/postgresql-94-rhel7

OpenShift Enterprise uses Software Collections (SCLs) to install and launch PostgreSQL. If you want to execute a PostgreSQL command inside of a running container (for debugging), you must invoke it using bash.

To do so, first identify the name of the running PostgreSQL pod. For example, you can view the list of pods in your current project:

$ oc get pods

Then, open a remote shell session to the desired pod:

$ oc rsh <pod>

When you enter the container, the required SCL is automatically enabled.

You can now run the psql command from the bash shell to start a PostgreSQL interactive session and perform normal PostgreSQL operations. For example, to authenticate as the database user:

bash-4.2$ PGPASSWORD=$POSTGRESQL_PASSWORD psql -h postgresql $POSTGRESQL_DATABASE $POSTGRESQL_USER
psql (9.2.8)
Type "help" for help.

default=>

When you are finished, enter \q to leave the PostgreSQL session.

The PostgreSQL user name, password, and database name must be configured with the following environment variables:

PostgreSQL settings can be configured with the following environment variables:

The PostgreSQL image can be run with mounted volumes to enable persistent storage for the database:

Passwords are part of the image configuration, therefore the only supported method to change passwords for the database user (POSTGRESQL_USER) and postgres administrator user is by changing the environment variables POSTGRESQL_PASSWORD and POSTGRESQL_ADMIN_PASSWORD, respectively.

You can view the current passwords by viewing the pod or deployment configuration in the web console or by listing the environment variables with the CLI:

$ oc set env pod <pod_name> --list

Changing database passwords through SQL statements or any way other than through the environment variables aforementioned will cause a mismatch between the values stored in the variables and the actual passwords. Whenever a database container starts, it resets the passwords to the values stored in the environment variables.

To change these passwords, update one or both of the desired environment variables for the related deployment configuration(s) using the oc set env command. If multiple deployment configurations utilize these environment variables, for example in the case of an application created from a template, you must update the variables on each deployment configuration so that the passwords are in sync everywhere. This can be done all in the same command:

$ oc set env dc <dc_name> [<dc_name_2> ...] \
  POSTGRESQL_PASSWORD=<new_password> \
  POSTGRESQL_ADMIN_PASSWORD=<new_admin_password>

Updating the environment variables triggers the redeployment of the database server if you have a configuration change trigger. Otherwise, you must manually start a new deployment in order to apply the password changes.

To verify that new passwords are in effect, first open a remote shell session to the running PostgreSQL pod:

$ oc rsh <pod>

From the bash shell, verify the database user’s new password:

bash-4.2$ PGPASSWORD=<new_password> psql -h postgresql $POSTGRESQL_DATABASE $POSTGRESQL_USER -c "SELECT * FROM (SELECT current_database()) cdb CROSS JOIN (SELECT current_user) cu"

If the password was changed correctly, you should see a table like this:

 current_database | current_user
------------------+--------------
 default          | django
(1 row)

From the bash shell, verify the postgres administrator user’s new password:

bash-4.2$ PGPASSWORD=<new_admin_password> psql -h postgresql $POSTGRESQL_DATABASE postgres -c "SELECT * FROM (SELECT current_database()) cdb CROSS JOIN (SELECT current_user) cu"

If the password was changed correctly, you should see a table like this:

 current_database | current_user
------------------+--------------
 default          | postgres
(1 row)

You can configure MongoDB with an ephemeral volume or a persistent volume. The first time you use the volume, the database is created along with the database administrator user. Afterwards, the MongoDB daemon starts up. If you are re-attaching the volume to another container, then the database, database user, and the administrator user are not created, and the MongoDB daemon starts.

The following command creates a new database pod with MongoDB running in a container with an ephemeral volume:

$ oc new-app \
    -e MONGODB_USER=<username> \
    -e MONGODB_PASSWORD=<password> \
    -e MONGODB_DATABASE=<database_name> \
    -e MONGODB_ADMIN_PASSWORD=<admin_password> \
    registry.access.redhat.com/rhscl/mongodb-26-rhel7

OpenShift Enterprise uses Software Collections (SCLs) to install and launch MongoDB. If you want to execute a MongoDB command inside of a running container (for debugging), you must invoke it using bash.

To do so, first identify the name of the running MongoDB pod. For example, you can view the list of pods in your current project:

$ oc get pods

Then, open a remote shell session to the desired pod:

$ oc rsh <pod>

When you enter the container, the required SCL is automatically enabled.

You can now run mongo commands from the bash shell to start a MongoDB interactive session and perform normal MongoDB operations. For example, to switch to the sampledb database and authenticate as the database user:

bash-4.2$ mongo -u $MONGODB_USER -p $MONGODB_PASSWORD $MONGODB_DATABASE
MongoDB shell version: 2.4.9
connecting to: sampledb
>

When you are finished, press CTRL+D to leave the MongoDB session.

The MongoDB user name, password, database name, and admin password must be configured with the following environment variables:

MongoDB settings can be configured with the following environment variables:

The MongoDB image can be run with mounted volumes to enable persistent storage for the database:

Passwords are part of the image configuration, therefore the only supported method to change passwords for the database user (MONGODB_USER) and admin user is by changing the environment variables MONGODB_PASSWORD and MONGODB_ADMIN_PASSWORD, respectively.

You can view the current passwords by viewing the pod or deployment configuration in the web console or by listing the environment variables with the CLI:

$ oc set env pod <pod_name> --list

Changing database passwords directly in MongoDB causes a mismatch between the values stored in the variables and the actual passwords. Whenever a database container starts, it resets the passwords to the values stored in the environment variables.

To change these passwords, update one or both of the desired environment variables for the related deployment configuration(s) using the oc set env command. If multiple deployment configurations utilize these environment variables, for example in the case of an application created from a template, you must update the variables on each deployment configuration so that the passwords are in sync everywhere. This can be done all in the same command:

$ oc set env dc <dc_name> [<dc_name_2> ...] \
  MONGODB_PASSWORD=<new_password> \
  MONGODB_ADMIN_PASSWORD=<new_admin_password>

Updating the environment variables triggers the redeployment of the database server if you have a configuration change trigger. Otherwise, you must manually start a new deployment in order to apply the password changes.

To verify that new passwords are in effect, first open a remote shell session to the running MongoDB pod:

$ oc rsh <pod>

From the bash shell, verify the database user’s new password:

bash-4.2$ mongo -u $MONGODB_USER -p <new_password> $MONGODB_DATABASE --eval "db.version()"

If the password was changed correctly, you should see output like this:

MongoDB shell version: 2.6.9
connecting to: sampledb
2.6.9

To verify the admin user’s new password:

bash-4.2$ mongo -u admin -p <new_admin_password> admin --eval "db.version()"

If the password was changed correctly, you should see output like this:

MongoDB shell version: 2.4.9
connecting to: admin
2.4.9

To set up MongoDB replication, a deployment configuration is defined in the example template that defines a replication controller. The replication controller manages the members of the MongoDB cluster.

To tell a MongoDB server that the member will be part of the cluster, additional environment variables are provided for the container defined in the replication controller pod template:

kind: DeploymentConfig
apiVersion: v1
metadata:
  name: "${MONGODB_SERVICE_NAME}"
spec:
  strategy:
    type: Recreate
    resources: {}
  triggers:
    - type: ConfigChange
  replicas: 3
  selector:
    name: mongodb-replica
  template:
    metadata:
      labels:
        name: mongodb-replica
    spec:
      containers:
        - name: member
          image: openshift/mongodb-24-centos7
          env:
            - name: MONGODB_USER
              value: "${MONGODB_USER}"
            - name: MONGODB_PASSWORD
              value: "${MONGODB_PASSWORD}"
            - name: MONGODB_DATABASE
              value: "${MONGODB_DATABASE}"
            - name: MONGODB_ADMIN_PASSWORD
              value: "${MONGODB_ADMIN_PASSWORD}"
            - name: MONGODB_REPLICA_NAME
              value: "${MONGODB_REPLICA_NAME}"
            - name: MONGODB_SERVICE_NAME
              value: "${MONGODB_SERVICE_NAME}"
            - name: MONGODB_KEYFILE_VALUE
              value: "${MONGODB_KEYFILE_VALUE}"
          ports:
            - containerPort: 27017
              protocol: TCP
  restartPolicy: Never
  dnsPolicy: ClusterFirst

After the deployment configuration is created and the pods with MongoDB cluster members are started, they will not be initialized. Instead, they start as part of the rs0 replication set, as the value of MONGODB_REPLICA_NAME is set to rs0 by default.

To initialize members created by the deployment configuration, the pods are started with the initiate argument, which instructs the startup script to behave slightly differently than a regular, stand-alone MongoDB database.

- kind: DeploymentConfig
  apiVersion: v1
  metadata:
    name: "${MONGODB_SERVICE_NAME}"
  spec:
    strategy:
      type: Recreate
      recreateParams:
        post:
          failurePolicy: Retry
          execNewPod:
            command: ["run-mongod","initiate"]
            containerName: mongodb
            env:
            - name: MONGODB_INITIAL_REPLICA_COUNT
              value: '3'
    triggers:
    - type: ConfigChange
    replicas: 3
    selector:
      name: mongodb-replica
    template:
      metadata:
        labels:
          name: mongodb-replica
      spec:
        containers:
        - name: mongodb
          image: openshift/mongodb-24-centos7
          readinessProbe:
            tcpSocket:
              port: 27017
            initialDelaySeconds: 15
            timeoutSeconds: 1
          env:
          - name: MONGODB_USER
            value: "${MONGODB_USER}"
          - name: MONGODB_PASSWORD
            value: "${MONGODB_PASSWORD}"
          - name: MONGODB_DATABASE
            value: "${MONGODB_DATABASE}"
          - name: MONGODB_ADMIN_PASSWORD
            value: "${MONGODB_ADMIN_PASSWORD}"
          - name: MONGODB_REPLICA_NAME
            value: "${MONGODB_REPLICA_NAME}"
          - name: MONGODB_SERVICE_NAME
            value: "${MONGODB_SERVICE_NAME}"
          - name: MONGODB_KEYFILE_VALUE
            value: "${MONGODB_KEYFILE_VALUE}"
          ports:
          - containerPort: 27017

The initiate argument in the container specification above instructs the container to first discover all running member pods within the MongoDB cluster. To achieve this, a headless service is defined named mongodb in the example template.

To have a headless service, the portalIP parameter in the service definition is set to None. Then you can use a DNS query to get a list of the pod IP addresses that represents the current endpoints for this service.

kind: "Service"
apiVersion: "v1"
metadata:
  name: "${MONGODB_SERVICE_NAME}"
  labels:
    name: "${MONGODB_SERVICE_NAME}"
spec:
  ports:
    - protocol: "TCP"
      port: 27017
      targetPort: 27017
      nodePort: 0
  selector:
    name: "mongodb-replica"
  portalIP: "None"
  type: "ClusterIP"
  sessionAffinity: "None"
status:
  loadBalancer: {}

When the script that runs as the container entrypoint has the IP addresses of all running MongoDB members, it creates a MongoDB replication set configuration where it lists all member IP addresses. It then initiates the replication set using rs.initiate(config). The script waits until MongoDB elects the PRIMARY member of the cluster.

Once the PRIMARY member has been elected, the entrypoint script starts creating MongoDB users and databases.

Clients can then start using the MongoDB instance by sending the queries to the mongodb service. As this service is a headless service, they do not need to provide the IP address. Clients can use mongodb:27017 for connections. The service then sends the query to one of the members in the replication set.

To increase the number of members in the cluster:

$ oc scale rc mongodb-1 --replicas=<number>

This tells the replication controller to create a new MongoDB member pod. When a new member is created, the member entrypoint first attempts to discover other running members in the cluster. It then chooses one and adds itself to the list of members. Once the replication configuration is updated, the other members replicate the data to a new pod and start a new election.