Add SqlJob and Flink Streaming impl (#69)

linkedin · Aug 12, 2024 · a400db2 · a400db2
1 parent 701ce75
commit a400db2
Show file tree

Hide file tree

Showing 31 changed files with 1,300 additions and 122 deletions.
diff --git a/Makefile b/Makefile
@@ -34,7 +34,8 @@ deploy-dev-environment:
 deploy-samples: deploy
 	kubectl wait --for=condition=Established=True	\
 	  crds/subscriptions.hoptimator.linkedin.com \
-	  crds/kafkatopics.hoptimator.linkedin.com
+	  crds/kafkatopics.hoptimator.linkedin.com \
+	  crds/sqljobs.hoptimator.linkedin.com
 	kubectl apply -f ./deploy/samples
 
 deploy-config:

diff --git a/deploy/rbac.yaml b/deploy/rbac.yaml
@@ -5,10 +5,10 @@ metadata:
   name: hoptimator-operator
 rules:
 - apiGroups: ["hoptimator.linkedin.com"]
-  resources: ["acls", "kafkatopics", "subscriptions"]
+  resources: ["acls", "kafkatopics", "subscriptions", "sqljobs"]
   verbs: ["get", "watch", "list", "update", "create"]
 - apiGroups: ["hoptimator.linkedin.com"]
-  resources: ["kafkatopics/status", "subscriptions/status", "acls/status"]
+  resources: ["kafkatopics/status", "subscriptions/status", "acls/status", "sqljobs/status"]
   verbs: ["get", "patch"]
 - apiGroups: ["flink.apache.org"]
   resources: ["flinkdeployments"]

diff --git a/deploy/samples/sqljobs.yaml b/deploy/samples/sqljobs.yaml
@@ -0,0 +1,10 @@
+apiVersion: hoptimator.linkedin.com/v1alpha1
+kind: SqlJob
+metadata:
+  name: hello-world
+spec:
+  dialect: Flink
+  executionMode: Streaming
+  sql:
+  - create table bh (text varchar) with ('connector' = 'blackhole');
+  - insert into bh values ('hello world');
diff --git a/deploy/sqljobs.crd.yaml b/deploy/sqljobs.crd.yaml
@@ -0,0 +1,87 @@
+apiVersion: apiextensions.k8s.io/v1
+kind: CustomResourceDefinition
+metadata:
+  name: sqljobs.hoptimator.linkedin.com
+spec:
+  group: hoptimator.linkedin.com
+  names:
+    kind: SqlJob
+    listKind: SqlJobList
+    plural: sqljobs
+    singular: sqljob
+    shortNames:
+    - sql
+    - sj
+  preserveUnknownFields: false
+  scope: Namespaced
+  versions:
+    - name: v1alpha1
+      served: true
+      storage: true
+      schema:
+        openAPIV3Schema:
+          description: Hoptimator generic SQL job
+          type: object
+          properties:
+            apiVersion:
+              type: string
+            kind:
+              type: string
+            metadata:
+              type: object
+            spec:
+              description: SQL job spec
+              type: object
+              properties:
+                sql:
+                  description: SQL script the job should run.
+                  type: array
+                  items:
+                    type: string
+                dialect:
+                  description: Flink, etc.
+                  type: string
+                  enum:
+                  - Flink 
+                  default: Flink
+                executionMode:
+                  description: Streaming or Batch.
+                  type: string
+                  enum:
+                  - Streaming
+                  - Batch
+                  default: Streaming
+              required:
+              - sql
+            status:
+              description: Filled in by the operator.
+              type: object
+              properties:
+                ready:
+                  description: Whether the SqlJob is running or completed.
+                  type: boolean
+                failed:
+                  description: Whether the SqlJob has failed.
+                  type: boolean
+                message:
+                  description: Error or success message, for information only.
+                  type: string
+                sql:
+                  description: The SQL being implemented by this SqlJob.
+                  type: string
+      subresources:
+        status: {}
+      additionalPrinterColumns:
+      - name: DIALECT
+        type: string
+        description: SQL dialect.
+        jsonPath: .spec.dialect
+      - name: MODE
+        type: string
+        description: Execution mode.
+        jsonPath: .spec.executionMode
+      - name: STATUS
+        type: string
+        description: Job status.
+        jsonPath: .status.message
+
diff --git a/hoptimator-flink-adapter/build.gradle b/hoptimator-flink-adapter/build.gradle
@@ -6,6 +6,7 @@ plugins {
 
 dependencies {
   implementation project(':hoptimator-catalog')
+  implementation project(':hoptimator-models')
   implementation project(':hoptimator-operator')
   implementation libs.kubernetesClient
   implementation libs.kubernetesExtendedClient

diff --git a/...ink-adapter/src/main/java/com/linkedin/hoptimator/catalog/flink/FlinkStreamingSqlJob.java b/...ink-adapter/src/main/java/com/linkedin/hoptimator/catalog/flink/FlinkStreamingSqlJob.java
@@ -0,0 +1,13 @@
+package com.linkedin.hoptimator.catalog.flink;
+
+import com.linkedin.hoptimator.catalog.Resource;
+
+public class FlinkStreamingSqlJob extends Resource {
+
+  public FlinkStreamingSqlJob(String namespace, String name, String sql) {
+    super("FlinkStreamingSqlJob");
+    export("namespace", namespace);
+    export("name", name);
+    export("sql", sql);
+  }
+}
diff --git a/...adapter/src/main/java/com/linkedin/hoptimator/operator/flink/FlinkControllerProvider.java b/...adapter/src/main/java/com/linkedin/hoptimator/operator/flink/FlinkControllerProvider.java
@@ -1,9 +1,13 @@
 package com.linkedin.hoptimator.operator.flink;
 
+import com.linkedin.hoptimator.models.V1alpha1SqlJob;
+import com.linkedin.hoptimator.models.V1alpha1SqlJobList;
 import com.linkedin.hoptimator.operator.ControllerProvider;
 import com.linkedin.hoptimator.operator.Operator;
 
 import io.kubernetes.client.extended.controller.Controller;
+import io.kubernetes.client.extended.controller.builder.ControllerBuilder;
+import io.kubernetes.client.extended.controller.reconciler.Reconciler;
 
 import java.util.Collection;
 import java.util.Collections;
@@ -16,7 +20,17 @@ public Collection<Controller> controllers(Operator operator) {
     operator.registerApi("FlinkDeployment", "flinkdeployment", "flinkdeployments",
       "flink.apache.org", "v1beta1");
 
-    // We don't need a controller
-    return Collections.emptyList();
+    operator.registerApi("SqlJob", "sqljob", "sqljobs",
+      "hoptimator.linkedin.com", "v1alpha1", V1alpha1SqlJob.class, V1alpha1SqlJobList.class);
+
+    Reconciler reconciler = new FlinkStreamingSqlJobReconciler(operator);
+    Controller controller = ControllerBuilder.defaultBuilder(operator.informerFactory())
+      .withReconciler(reconciler)
+      .withName("flink-streaming-sql-job-controller")
+      .withWorkerCount(1)
+      .watch(x -> ControllerBuilder.controllerWatchBuilder(V1alpha1SqlJob.class, x).build())
+      .build();
+
+    return Collections.singleton(controller);
   }
 }
diff --git a/.../src/main/java/com/linkedin/hoptimator/operator/flink/FlinkStreamingSqlJobReconciler.java b/.../src/main/java/com/linkedin/hoptimator/operator/flink/FlinkStreamingSqlJobReconciler.java
@@ -0,0 +1,109 @@
+package com.linkedin.hoptimator.operator.flink;
+
+import com.linkedin.hoptimator.catalog.Resource;
+import com.linkedin.hoptimator.catalog.flink.FlinkStreamingSqlJob;
+import com.linkedin.hoptimator.operator.Operator;
+import com.linkedin.hoptimator.models.V1alpha1SqlJob;
+import com.linkedin.hoptimator.models.V1alpha1SqlJobSpec.DialectEnum;
+import com.linkedin.hoptimator.models.V1alpha1SqlJobSpec.ExecutionModeEnum;
+import com.linkedin.hoptimator.models.V1alpha1SqlJobStatus;
+
+import io.kubernetes.client.extended.controller.reconciler.Reconciler;
+import io.kubernetes.client.extended.controller.reconciler.Request;
+import io.kubernetes.client.extended.controller.reconciler.Result;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.util.List;
+import java.util.stream.Collectors;
+
+/**
+ * Manifests streaming SqlJobs as Flink jobs.
+ *
+ */
+public class FlinkStreamingSqlJobReconciler implements Reconciler {
+  private final static Logger log = LoggerFactory.getLogger(FlinkStreamingSqlJobReconciler.class);
+  private final static String SQLJOB = "hoptimator.linkedin.com/v1alpha1/SqlJob";
+
+  private final Operator operator;
+
+  public FlinkStreamingSqlJobReconciler(Operator operator) {
+    this.operator = operator;
+  }
+
+  @Override
+  public Result reconcile(Request request) {
+    log.info("Reconciling request {}", request);
+    String name = request.getName();
+    String namespace = request.getNamespace();
+    Result result = new Result(true, operator.pendingRetryDuration());
+
+    try {
+      V1alpha1SqlJob object = operator.<V1alpha1SqlJob>fetch(SQLJOB, namespace, name);
+
+      if (object == null) {
+        log.info("Object {}/{} deleted. Skipping.");
+        return new Result(false);
+      }
+
+      V1alpha1SqlJobStatus status = object.getStatus();
+      if (status == null) {
+        status = new V1alpha1SqlJobStatus();
+        object.setStatus(status);
+      }
+
+      List<String> sql = object.getSpec().getSql();
+      String script = sql.stream().collect(Collectors.joining(";\n"));
+
+      DialectEnum dialect = object.getSpec().getDialect();
+      if (!DialectEnum.FLINK.equals(dialect)) {
+        log.info("Not Flink SQL. Skipping.");
+        return new Result(false);
+      }
+
+      ExecutionModeEnum mode = object.getSpec().getExecutionMode();
+      if (!ExecutionModeEnum.STREAMING.equals(mode)) {
+        log.info("Not a streaming job. Skipping.");
+        return new Result(false);
+      }
+
+      Resource.TemplateFactory templateFactory = new Resource.SimpleTemplateFactory(Resource.Environment.EMPTY);
+      Resource sqlJob = new FlinkStreamingSqlJob(namespace, name, script);
+      boolean allReady = true;
+      boolean anyFailed = false;
+      for (String yaml : sqlJob.render(templateFactory)) {
+        operator.apply(yaml, object); 
+        if (!operator.isReady(yaml)) {
+          allReady = false;
+        }
+        if (operator.isFailed(yaml)) {
+          anyFailed = true;
+          allReady = false;
+        }
+      }
+
+      object.getStatus().setReady(allReady);
+      object.getStatus().setFailed(anyFailed);
+
+      if (allReady) {
+        object.getStatus().setMessage("Ready.");
+        result = new Result(false); // done
+      }
+      if (anyFailed) {
+        object.getStatus().setMessage("Failed.");
+        result = new Result(false); // done
+      }
+
+      operator.apiFor(SQLJOB).updateStatus(object, x -> object.getStatus())
+          .onFailure((x, y) -> log.error("Failed to update status of SqlJob {}: {}.", name, y.getMessage()))
+          .throwsApiException();
+
+    } catch (Exception e) {
+      log.error("Encountered exception while reconciling Flink streaming SqlJob {}/{}", namespace, name, e);
+      return new Result(true, operator.failureRetryDuration());
+    }
+    return result;
+  }
+}
+
diff --git a/hoptimator-flink-adapter/src/main/resources/FlinkStreamingSqlJob.yaml.template b/hoptimator-flink-adapter/src/main/resources/FlinkStreamingSqlJob.yaml.template
@@ -0,0 +1,29 @@
+apiVersion: flink.apache.org/v1beta1
+kind: FlinkDeployment
+metadata:
+  namespace: {{namespace}}
+  name: {{name}}-flink-job
+spec:
+  image: docker.io/library/hoptimator-flink-runner
+  imagePullPolicy: Never
+  flinkVersion: v1_16
+  flinkConfiguration:
+    taskmanager.numberOfTaskSlots: "1"
+  serviceAccount: flink
+  jobManager:
+    resource:
+      memory: "2048m"
+      cpu: 0.1
+  taskManager:
+    resource:
+      memory: "2048m"
+      cpu: 0.1
+  job:
+    entryClass: com.linkedin.hoptimator.flink.runner.FlinkRunner
+    args:
+    - {{sql}}
+    jarURI: local:///opt/hoptimator-flink-runner.jar
+    parallelism: 1
+    upgradeMode: stateless
+    state: running
+
diff --git a/hoptimator-models/generate-models.sh b/hoptimator-models/generate-models.sh
@@ -13,4 +13,5 @@ docker run \
   -u "$(pwd)/deploy/acls.crd.yaml" \
   -u "$(pwd)/deploy/kafkatopics.crd.yaml" \
   -u "$(pwd)/deploy/subscriptions.crd.yaml" \
+  -u "$(pwd)/deploy/sqljobs.crd.yaml" \
   && echo "done."
diff --git a/hoptimator-models/src/main/java/com/linkedin/hoptimator/models/V1alpha1Acl.java b/hoptimator-models/src/main/java/com/linkedin/hoptimator/models/V1alpha1Acl.java
@@ -31,7 +31,7 @@
  * Access control rule (colloquially, an Acl)
  */
 @ApiModel(description = "Access control rule (colloquially, an Acl)")
-@javax.annotation.Generated(value = "org.openapitools.codegen.languages.JavaClientCodegen", date = "2024-02-09T06:53:38.470Z[Etc/UTC]")
+@javax.annotation.Generated(value = "org.openapitools.codegen.languages.JavaClientCodegen", date = "2024-07-26T20:21:01.735Z[Etc/UTC]")
 public class V1alpha1Acl implements io.kubernetes.client.common.KubernetesObject {
   public static final String SERIALIZED_NAME_API_VERSION = "apiVersion";
   @SerializedName(SERIALIZED_NAME_API_VERSION)

diff --git a/hoptimator-models/src/main/java/com/linkedin/hoptimator/models/V1alpha1AclList.java b/hoptimator-models/src/main/java/com/linkedin/hoptimator/models/V1alpha1AclList.java
@@ -32,7 +32,7 @@
  * AclList is a list of Acl
  */
 @ApiModel(description = "AclList is a list of Acl")
-@javax.annotation.Generated(value = "org.openapitools.codegen.languages.JavaClientCodegen", date = "2024-02-09T06:53:38.470Z[Etc/UTC]")
+@javax.annotation.Generated(value = "org.openapitools.codegen.languages.JavaClientCodegen", date = "2024-07-26T20:21:01.735Z[Etc/UTC]")
 public class V1alpha1AclList implements io.kubernetes.client.common.KubernetesListObject {
   public static final String SERIALIZED_NAME_API_VERSION = "apiVersion";
   @SerializedName(SERIALIZED_NAME_API_VERSION)

diff --git a/hoptimator-models/src/main/java/com/linkedin/hoptimator/models/V1alpha1AclSpec.java b/hoptimator-models/src/main/java/com/linkedin/hoptimator/models/V1alpha1AclSpec.java
@@ -29,7 +29,7 @@
  * A set of related ACL rules.
  */
 @ApiModel(description = "A set of related ACL rules.")
-@javax.annotation.Generated(value = "org.openapitools.codegen.languages.JavaClientCodegen", date = "2024-02-09T06:53:38.470Z[Etc/UTC]")
+@javax.annotation.Generated(value = "org.openapitools.codegen.languages.JavaClientCodegen", date = "2024-07-26T20:21:01.735Z[Etc/UTC]")
 public class V1alpha1AclSpec {
   /**
    * The resource access method.

diff --git a/hoptimator-models/src/main/java/com/linkedin/hoptimator/models/V1alpha1AclSpecResource.java b/hoptimator-models/src/main/java/com/linkedin/hoptimator/models/V1alpha1AclSpecResource.java
@@ -28,7 +28,7 @@
  * The resource being controlled.
  */
 @ApiModel(description = "The resource being controlled.")
-@javax.annotation.Generated(value = "org.openapitools.codegen.languages.JavaClientCodegen", date = "2024-02-09T06:53:38.470Z[Etc/UTC]")
+@javax.annotation.Generated(value = "org.openapitools.codegen.languages.JavaClientCodegen", date = "2024-07-26T20:21:01.735Z[Etc/UTC]")
 public class V1alpha1AclSpecResource {
   public static final String SERIALIZED_NAME_KIND = "kind";
   @SerializedName(SERIALIZED_NAME_KIND)

diff --git a/hoptimator-models/src/main/java/com/linkedin/hoptimator/models/V1alpha1AclStatus.java b/hoptimator-models/src/main/java/com/linkedin/hoptimator/models/V1alpha1AclStatus.java
@@ -28,7 +28,7 @@
  * Status, as set by the operator.
  */
 @ApiModel(description = "Status, as set by the operator.")
-@javax.annotation.Generated(value = "org.openapitools.codegen.languages.JavaClientCodegen", date = "2024-02-09T06:53:38.470Z[Etc/UTC]")
+@javax.annotation.Generated(value = "org.openapitools.codegen.languages.JavaClientCodegen", date = "2024-07-26T20:21:01.735Z[Etc/UTC]")
 public class V1alpha1AclStatus {
   public static final String SERIALIZED_NAME_MESSAGE = "message";
   @SerializedName(SERIALIZED_NAME_MESSAGE)

diff --git a/hoptimator-models/src/main/java/com/linkedin/hoptimator/models/V1alpha1KafkaTopic.java b/hoptimator-models/src/main/java/com/linkedin/hoptimator/models/V1alpha1KafkaTopic.java
@@ -31,7 +31,7 @@
  * Kafka Topic
  */
 @ApiModel(description = "Kafka Topic")
-@javax.annotation.Generated(value = "org.openapitools.codegen.languages.JavaClientCodegen", date = "2024-02-09T06:53:38.470Z[Etc/UTC]")
+@javax.annotation.Generated(value = "org.openapitools.codegen.languages.JavaClientCodegen", date = "2024-07-26T20:21:01.735Z[Etc/UTC]")
 public class V1alpha1KafkaTopic implements io.kubernetes.client.common.KubernetesObject {
   public static final String SERIALIZED_NAME_API_VERSION = "apiVersion";
   @SerializedName(SERIALIZED_NAME_API_VERSION)

diff --git a/hoptimator-models/src/main/java/com/linkedin/hoptimator/models/V1alpha1KafkaTopicList.java b/hoptimator-models/src/main/java/com/linkedin/hoptimator/models/V1alpha1KafkaTopicList.java
@@ -32,7 +32,7 @@
  * KafkaTopicList is a list of KafkaTopic
  */
 @ApiModel(description = "KafkaTopicList is a list of KafkaTopic")
-@javax.annotation.Generated(value = "org.openapitools.codegen.languages.JavaClientCodegen", date = "2024-02-09T06:53:38.470Z[Etc/UTC]")
+@javax.annotation.Generated(value = "org.openapitools.codegen.languages.JavaClientCodegen", date = "2024-07-26T20:21:01.735Z[Etc/UTC]")
 public class V1alpha1KafkaTopicList implements io.kubernetes.client.common.KubernetesListObject {
   public static final String SERIALIZED_NAME_API_VERSION = "apiVersion";
   @SerializedName(SERIALIZED_NAME_API_VERSION)

diff --git a/hoptimator-models/src/main/java/com/linkedin/hoptimator/models/V1alpha1KafkaTopicSpec.java b/hoptimator-models/src/main/java/com/linkedin/hoptimator/models/V1alpha1KafkaTopicSpec.java
@@ -33,7 +33,7 @@
  * Desired Kafka topic configuration.
  */
 @ApiModel(description = "Desired Kafka topic configuration.")
-@javax.annotation.Generated(value = "org.openapitools.codegen.languages.JavaClientCodegen", date = "2024-02-09T06:53:38.470Z[Etc/UTC]")
+@javax.annotation.Generated(value = "org.openapitools.codegen.languages.JavaClientCodegen", date = "2024-07-26T20:21:01.735Z[Etc/UTC]")
 public class V1alpha1KafkaTopicSpec {
   public static final String SERIALIZED_NAME_CLIENT_CONFIGS = "clientConfigs";
   @SerializedName(SERIALIZED_NAME_CLIENT_CONFIGS)