remove isRunningOnEMR

regex is specific enough to avoid collision

Author: aboitreaud

dd-java-agent/instrumentation/spark/spark-common/build.gradle

@@ -28,7 +28,6 @@ dependencies {
   testFixturesCompileOnly(libs.bundles.spock)
 
   testImplementation project(':dd-java-agent:instrumentation-testing')
-  testImplementation group: 'org.apache.spark', name: 'spark-core_2.12', version: '2.4.0'
   testImplementation group: 'org.apache.spark', name: 'spark-launcher_2.12', version: '2.4.0'
 }
 

dd-java-agent/instrumentation/spark/spark-common/src/main/java/datadog/trace/instrumentation/spark/AbstractDatadogSparkListener.java

@@ -134,7 +134,6 @@ public abstract class AbstractDatadogSparkListener extends SparkListener {
 
   private volatile boolean isStreamingJob = false;
   private final boolean isRunningOnDatabricks;
-  private final boolean isRunningOnEmr;
   private final String databricksClusterName;
   private final String databricksServiceName;
   private final String sparkServiceName;
@@ -157,7 +156,6 @@ public AbstractDatadogSparkListener(SparkConf sparkConf, String appId, String sp
     this.sparkVersion = sparkVersion;
 
     isRunningOnDatabricks = sparkConf.contains("spark.databricks.sparkContextId");
-    isRunningOnEmr = EmrUtils.isRunningOnEmr(sparkConf);
     databricksClusterName = sparkConf.get("spark.databricks.clusterUsageTags.clusterName", null);
     databricksServiceName = getDatabricksServiceName(sparkConf, databricksClusterName);
     sparkServiceName = getSparkServiceName(sparkConf, isRunningOnDatabricks);
@@ -278,7 +276,7 @@ private void initApplicationSpanIfNotInitialized() {
     }
 
     captureApplicationParameters(builder);
-    captureEmrStepId(builder, isRunningOnEmr);
+    captureEmrStepId(builder);
 
     Optional<OpenlineageParentContext> openlineageParentContext =
         OpenlineageParentContext.from(sparkConf);
@@ -1213,10 +1211,7 @@ private void captureApplicationParameters(AgentTracer.SpanBuilder builder) {
     builder.withTag("config.spark_version", sparkVersion);
   }
 
-  private static void captureEmrStepId(AgentTracer.SpanBuilder builder, boolean isRunningOnEmr) {
-    if (!isRunningOnEmr) {
-      return;
-    }
+  private static void captureEmrStepId(AgentTracer.SpanBuilder builder) {
     String stepId = EmrUtils.getEmrStepId();
     if (stepId != null) {
       builder.withTag("emr_step_id", stepId);

dd-java-agent/instrumentation/spark/spark-common/src/main/java/datadog/trace/instrumentation/spark/EmrUtils.java

@@ -2,32 +2,20 @@
 
 import java.nio.file.Path;
 import java.nio.file.Paths;
-import java.util.Arrays;
 import java.util.regex.Matcher;
 import java.util.regex.Pattern;
 import javax.annotation.Nullable;
-import org.apache.spark.SparkConf;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
-/** Utilities for detecting AWS EMR and extracting EMR-specific metadata. */
+/** Extracts the AWS EMR Step ID from the working directory name (e.g. s-07767992IY7VC5NVV854). */
 class EmrUtils {
 
   private static final Logger log = LoggerFactory.getLogger(EmrUtils.class);
 
   /** EMR step ID is a 20 character string with numbers and uppercase letters only */
   private static final Pattern EMR_STEP_ID_PATTERN = Pattern.compile("^(s-[0-9A-Z]{20})$");
 
-  /**
-   * Returns true if the Spark job is running on AWS EMR. Detection scans all SparkConf keys for the
-   * {@code .emr.} substring, which is present in EMR-specific keys (e.g. {@code
-   * spark.emr.default.executor.cores}, {@code spark.sql.emr.internal.extensions}) and absent from
-   * all standard Spark configuration keys.
-   */
-  static boolean isRunningOnEmr(SparkConf conf) {
-    return Arrays.stream(conf.getAll()).anyMatch(t -> t._1().contains(".emr."));
-  }
-
   @Nullable
   static String getEmrStepId() {
     try {

dd-java-agent/instrumentation/spark/spark-common/src/test/java/datadog/trace/instrumentation/spark/EmrUtilsTest.java

@@ -2,9 +2,7 @@
 
 import static org.junit.jupiter.api.Assertions.assertEquals;
 import static org.junit.jupiter.api.Assertions.assertNull;
-import static org.junit.jupiter.api.Assertions.assertTrue;
 
-import org.apache.spark.SparkConf;
 import org.junit.jupiter.api.AfterEach;
 import org.junit.jupiter.api.BeforeEach;
 import org.junit.jupiter.api.Test;
@@ -40,12 +38,4 @@ void returnsNullForApplicationIdWorkdir() {
     System.setProperty("user.dir", "/home/hadoop/application_1234567890_0001");
     assertNull(EmrUtils.getEmrStepId());
   }
-
-  @Test
-  void isRunningOnEmrReturnsTrueWhenConfContainsEmrKey() {
-    SparkConf conf = new SparkConf();
-    conf.set(
-        "spark.sql.emr.internal.extensions", "com.amazonaws.emr.spark.EmrSparkSessionExtensions");
-    assertTrue(EmrUtils.isRunningOnEmr(conf));
-  }
 }