Deploying jobs by using the Apache SQL Runner sample

Find out how to deploy your flows in an application mode Flink cluster as part of your production environment.

Important:

You can use the JSON and configuration YAML flow export format for deploying jobs customized for production or test environments. In most cases, this provides a better user-experience, and can be used with an automation in a continuous integration and continuous delivery (CI/CD) pipeline.

This deployment cannot be used with the Event Processing UI.

Note: The Apache operator sample that is referenced in the following sections points to the version of the sample in the main branch, which is up-to-date, and might include fixes that are absent in the release branches.

Prerequisites

The SQL statements are exported from the Event Processing UI and saved to a file, for example, statements.sql.

For more information, see exporting flows.

You updated the Flink SQL connector properties and values that are defined in the file statements.sql to match your target environment.

For security reasons, the values containing sensitive credentials, such as username and password, are removed when exporting the SQL statements from the Event Processing UI, so you must restore them.

Also, the exported SQL contains connector configuration that is applicable to the environment targeted in the Event Processing UI. When deploying to a different target environment, you might need to adapt the connector properties to the target environment.

See the following table for credentials and connector properties information about supported Flink SQL connectors:

Flink SQL connector	Used by nodes	Sensitive credentials	Connector properties
Kafka	Source and destination	About Kafka connector Note: When configuring SCRAM authentication for the Kafka connector, ensure you use double quotes only. Do not use a backslash character (`\`) to escape the double quotes. The valid format is: `username="<username>" password="<password>"`	Kafka connector properties For more information about how events can be consumed from Kafka topics, see the Flink documentation. Note: The Kafka connector value must be `kafka`. The Kafka connector version for Event Processing versions 1.3.0 and later is 3.4.0-1.20.
JDBC	Database	About JDBC connector	JDBC connector properties
HTTP	API	About HTTP connector	HTTP connector properties. Note: The HTTP connector version for Event Processing versions 1.3.0 and later is 0.18.0.

To deploy a running Flink job, the SQL statements in the file statements.sql must contain one of the following clauses:
- A definition of a Flink SQL Kafka sink (also known as event destination), and an INSERT INTO clause that selects the columns of the last temporary view into this sink.
- A SELECT clause that takes one or all of the columns of the last temporary view.
For more information about how to define a Flink SQL sink, see the Flink documentation.

Use Flink user-defined functions

Optionally, user-defined functions (UDFs) can be used as a complement of the built-in functions, by editing the SQL exported from the Event Processing UI.

For more information, see UDFs in the exported SQL.

Setup a connection to the Flink cluster

Log in to your Kubernetes cluster as a cluster administrator by setting your kubectl context.
Switch to the namespace where the IBM Operator for Apache Flink is installed:
```
kubectl config set-context --current --namespace=<namespace>
```

Build and deploy a Flink SQL runner

You can use a Kubernetes FlinkDeployment custom resource in application mode to deploy a Flink job for processing and deploying the statements in the file statements.sql.

A sample application flink-sql-runner-example is provided in the Apache Flink GitHub repository for that purpose.

Follow the instructions to build:

the flink-sql-runner-example JAR file (Flink job)
the Docker image

Important: Ensure that the Flink SQL runner JAR file and the statements.sql file have read permissions (644) for non-root users. If the JAR file is only readable by the root user, the FlinkDeployment instance cannot be started by non-root users.

Some adaptations to this procedure are required to build the Docker image and use the file statements.sql:

Modify the Dockerfile to use the IBM Flink image:

a. Execute the following command to extract the Flink image name including its SHA digest from the ClusterServiceVersion (CSV). For example, if you are running on Flink version 1.3.0:
```
kubectl get csv -o jsonpath='{.spec.install.spec.deployments[*].spec.template.spec.containers[0].env[?(@.name=="IBM_FLINK_IMAGE")].value}' ibm-eventautomation-flink.v1.3.0
```
Alternatively, you can obtain the image name from the Flink operator pod’s environment variable:
```
kubectl set env pod/<flink_operator_pod_name> --list -n <flink_operator_namespace> | grep IBM_FLINK_IMAGE
```
b. Edit the Dockerfile and change the FROM clause to use the IBM Flink image with its SHA digest, as determined in the previous step.
```
FROM --platform=<platform> <IBM Flink image with digest>
```
Where <platform> is linux/amd64 or linux/s390x, depending on your deployment target.

c. If at the previous optional step you introduced the use of Flink user-defined functions (UDFs), edit the Dockerfile to copy the JAR file that contains the UDF classes:
```
COPY --chown=flink:root <path-of-the-udf-jar> /opt/flink/lib
```
For example:
```
COPY --chown=flink:root /udfproject/target/udf.jar /opt/flink/lib
```
d. Remove the sample SQL statement files from the sql-scripts directory.

e. Copy the statements.sql file to the directory sql-scripts.

f. Build the docker image and push it to a registry accessible from your OpenShift Container Platform. If your registry requires authentication, configure the image pull secret, for example, by using the global cluster pull secret.

Create the IBM Operator for Apache Flink FlinkDeployment custom resource.

You can use a Kubernetes FlinkDeployment custom resource in application mode to deploy a Flink job for processing and deploying the statements in the statements.sql file.

a. You can start with the following example of an application mode Flink instance:

apiVersion: flink.apache.org/v1beta1
kind: FlinkDeployment
metadata:
name: application-cluster-prod
spec:
  image: <image built FROM icr.io/cpopen/ibm-eventautomation-flink/ibm-eventautomation-flink>
  flinkConfiguration:
    license.use: EventAutomationProduction
    license.license: 'L-KCVZ-JL5CRM'
    license.accept: 'false'
    high-availability.type: org.apache.flink.kubernetes.highavailability.KubernetesHaServicesFactory
    high-availability.storageDir: 'file:///opt/flink/volume/flink-ha'
    restart-strategy: failure-rate
    restart-strategy.failure-rate.max-failures-per-interval: '10'
    restart-strategy.failure-rate.failure-rate-interval: '10 min'
    restart-strategy.failure-rate.delay: '30 s'
    execution.checkpointing.interval: '5000'
    execution.checkpointing.unaligned: 'false'
    state.backend.type: rocksdb
    state.backend.rocksdb.thread.num: '10'
    state.backend.incremental: 'true'
    state.backend.rocksdb.use-bloom-filter: 'true'
    state.checkpoints.dir: 'file:///opt/flink/volume/flink-cp'
    state.checkpoints.num-retained: '3'
    state.savepoints.dir: 'file:///opt/flink/volume/flink-sp'
    taskmanager.numberOfTaskSlots: '2'
    table.exec.source.idle-timeout: '30 s'
    security.ssl.enabled: 'true'
    security.ssl.truststore: /opt/flink/tls-cert/truststore.jks
    security.ssl.truststore-password: <jks-password>
    security.ssl.keystore: /opt/flink/tls-cert/keystore.jks
    security.ssl.keystore-password: <jks-password>
    security.ssl.key-password: <jks-password>
    kubernetes.secrets: '<jks-secret>:/opt/flink/tls-cert'
serviceAccount: flink
podTemplate:
  apiVersion: v1
  kind: Pod
  metadata:
    name: pod-template
    spec:
      affinity:
        podAntiAffinity:
          preferredDuringSchedulingIgnoredDuringExecution:
            - weight: 80
              podAffinityTerm:
                labelSelector:
                  matchExpressions:
                    - key: type
                      operator: In
                      values:
                        - flink-native-kubernetes
                topologyKey: kubernetes.io/hostname
      containers:
        - name: flink-main-container
          volumeMounts:
            - name: flink-logs
              mountPath: /opt/flink/log
            - name: flink-volume
              mountPath: /opt/flink/volume
      volumes:
        - name: flink-logs
          emptyDir: {}
        - name: flink-volume
          persistentVolumeClaim:
            claimName: ibm-flink-pvc
jobManager:
  replicas: 2
  resource:
    memory: '4096m'
    cpu: 0.5
taskManager:
  resource:
    memory: '4096m'
    cpu: 2
job:
  jarURI: <insert jar file name here>
  args: ['<insert path for statements.sql here>']
  parallelism: 1
  state: running
  upgradeMode: savepoint
  allowNonRestoredState: true
mode: native

Note: The Flink instance must be configured with persistent storage.

If you do not want to use the examples provided earlier, add the following parameter to set a timeout period for event sources when they are marked idle. This allows downstream tasks to advance their watermark. Idleness is not detected by default. The parameter is included in all Flink samples:

spec:
flinkConfiguration:
   table.exec.source.idle-timeout: '30 s'

For more information about table.exec.source.idle-timeout, see the Flink documentation.

b. Append the following spec.job parameter, or edit the existing parameter if using the Production - Flink Application cluster sample:

spec:
  job:
    jarURI: local:///opt/flink/usrlib/sql-runner.jar
    args: ["/opt/flink/usrlib/sql-scripts/statements.sql"]
    parallelism: 1
    state: running
    upgradeMode: savepoint

c. Set the Flink image:

spec:
  image: <image built at step 1.e>

d. Set the Flink version (only required if the --set webhook.create is set to false during the operator installation).

Obtain the correct version by listing the IBM_FLINK_VERSION environment variable on the Flink operator pod:

kubectl set env pod/<flink_operator_pod_name> --list -n <flink_operator_namespace> | grep IBM_FLINK_VERSION

Add the version to the FlinkDeployment, for example:
```
spec:
  flinkVersion: "v1_19"
```

Apply the modified FlinkDeployment custom resource.

Changing the parallelism of a Flink SQL runner

Edit the FlinkDeployment custom resource.

a. Ensure that the Flink cluster has enough task slots to fulfill the targeted parallelism value.

Task slots = spec.taskmanager.replicas × spec.flinkConfiguration["taskmanager.numberOfTaskSlots"] 

b. Change the spec.job.parallelism value, then set spec.job.state to running and spec.job.upgradeMode to savepoint.

spec:
  job:
    jarURI: local:///opt/flink/usrlib/sql-runner.jar
    args: ["/opt/flink/usrlib/sql-scripts/statements.sql"]
    parallelism: 2
    state: running
    upgradeMode: savepoint
    allowNonRestoredState: true

Apply the modified FlinkDeployment custom resource.

The following operations are automatically performed by Flink:
- The FlinkStateSnapshots custom resource is triggered by the operator and a savepoint is created before the Flink job is suspended.
- The Flink cluster is shutdown, the JobManager and TaskManager pods are terminated.
- A Flink cluster is created with new JobManager and TaskManager pods.
- The Flink job is restarted from the savepoint.

Stop a Flink SQL job with a savepoint

You can temporarily stop a running Flink job while capturing its current state by creating a savepoint, and allowing you to restart the job from the exact point where it stopped by using the savepoint when required.

Edit the FlinkDeployment custom resource.

Make the following modifications:

a. Set that the value of spec.job.upgradeMode to savepoint.

b. Set that the value of spec.job.state to suspended to stop the Flink job.

spec:
  job:
    jarURI: local:///opt/flink/usrlib/sql-runner.jar
    args: ["/opt/flink/usrlib/sql-scripts/statements.sql"]
    state: suspended
    upgradeMode: savepoint

Save the changes in the FlinkDeployment custom resource.
The FlinkStateSnapshots custom resource is created by the operator with a name that starts with the name of your FlinkDeployment custom resource, and a savepoint is triggered and written to a location in the PVC, which is indicated in the status.path of FlinkStateSnapshots custom resource.

For example:
```
status:
  failures: 0
  path: 'file:/opt/flink/volume/flink-sp/savepoint-caf2b2-39d09a1c170c'
  state: COMPLETED
```

Resume a suspended Flink job

You can resume a suspended job from the exact point where it stopped by using the savepoint created during its suspension.

Edit the FlinkDeployment custom resource of a Flink job that you suspended earlier:

a. Set that the value of spec.job.upgradeMode is savepoint.

b. Set that the value of spec.job.state is running to resume the Flink job.

c. Set the value of spec.job.initialSavepointPath to the savepoint location found in the status.path field of the FlinkStateSnapshots custom resource from step 4 of stopping a Flink job with a savepoint.

For example:
```
job:
  jarURI: local:///opt/flink/usrlib/sql-runner.jar
  args: ["/opt/flink/usrlib/sql-scripts/statements.sql"]
  state: running
  upgradeMode: savepoint
  initialSavepointPath: file:/opt/flink/volume/flink-sp/savepoint-caf2b2-39d09a1c170c
  allowNonRestoredState: true
```
Save the changes in the FlinkDeployment custom resource.

For more information on manually restoring a job, see manual recovery.

Enable SSL connection for your database

To securely connect Flink jobs to a database such as PostgreSQL, MySQL, or Oracle, enable an SSL connection with the database as follows:

Ensure you added the CA certificate for your database to the truststore and then created a secret with the truststore.
Edit the FlinkDeployment custom resource.

Complete the following modifications:

In the spec.flinkConfiguration section, add:

env.java.opts.taskmanager: >-
   -Djavax.net.ssl.trustStore=/certs/truststore.<keystore-extension>
   -Djavax.net.ssl.trustStorePassword=<chosen password>
env.java.opts.jobmanager: >-
   -Djavax.net.ssl.trustStore=/certs/truststore.<keystore-extension>
   -Djavax.net.ssl.trustStorePassword=<chosen password>

Where:

<keystore-extension> is the extension for your keystore format. For example, jks for Java Keystore and p12 for Public-Key Cryptography Standards.

In spec.podTemplate.spec.containers.volumeMounts section, add:

- mountPath: /certs
  name: truststore
  readOnly: true

In spec.podTemplate.spec.volumes section, add:

- name: truststore
  secret:
    items:
      - key: truststore.<keystore-extension>
        path: truststore.<keystore-extension>
    secretName: ssl-truststore

Apply the modified FlinkDeployment custom resource:

kubectl apply -f <custom-resource-file-path>

For example:

kubectl apply -f flinkdeployment_demo.yaml

A secure SSL connection is enabled between Flink and the database.