Hive: Add Hive 4 support and remove Hive runtime

.github/labeler.yml

@@ -100,10 +100,7 @@ ORC:
 HIVE:
   - changed-files:
     - any-glob-to-any-file: [
-      'hive3/**/*',
       'hive-metastore/**/*',
-      'hive-runtime/**/*',
-      'hive3-orc-bundle/**/*'
     ]
 
 DATA:

.github/workflows/delta-conversion-ci.yml

@@ -48,9 +48,6 @@ on:
       - '.asf.yml'
       - 'dev/**'
       - 'mr/**'
-      - 'hive3/**'
-      - 'hive3-orc-bundle/**'
-      - 'hive-runtime/**'
       - 'flink/**'
       - 'kafka-connect/**'
       - 'docs/**'

.github/workflows/flink-ci.yml

@@ -48,9 +48,6 @@ on:
     - '.asf.yml'
     - 'dev/**'
     - 'mr/**'
-    - 'hive3/**'
-    - 'hive3-orc-bundle/**'
-    - 'hive-runtime/**'
     - 'kafka-connect/**'
     - 'spark/**'
     - 'docs/**'

.github/workflows/hive-ci.yml

@@ -87,39 +87,10 @@ jobs:
         key: ${{ runner.os }}-gradle-${{ hashFiles('**/*.gradle*', '**/gradle-wrapper.properties') }}
         restore-keys: ${{ runner.os }}-gradle-
     - run: echo -e "$(ip addr show eth0 | grep "inet\b" | awk '{print $2}' | cut -d/ -f1)\t$(hostname -f) $(hostname -s)" | sudo tee -a /etc/hosts
-    - run: ./gradlew -DsparkVersions= -DhiveVersions=2 -DflinkVersions= -DkafkaVersions= -Pquick=true :iceberg-mr:check :iceberg-hive-runtime:check -x javadoc
+    - run: ./gradlew -DsparkVersions= -DhiveVersions=2 -DflinkVersions= -DkafkaVersions= -Pquick=true :iceberg-mr:check -x javadoc
     - uses: actions/upload-artifact@v4
       if: failure()
       with:
         name: test logs
         path: |
-          **/build/testlogs
-
-  hive3-tests:
-    runs-on: ubuntu-22.04
-    strategy:
-      matrix:
-        jvm: [11, 17, 21]
-    env:
-      SPARK_LOCAL_IP: localhost
-    steps:
-    - uses: actions/checkout@v4
-    - uses: actions/setup-java@v4
-      with:
-        distribution: zulu
-        java-version: ${{ matrix.jvm }}
-    - uses: actions/cache@v4
-      with:
-        path: |
-          ~/.gradle/caches
-          ~/.gradle/wrapper
-        key: ${{ runner.os }}-gradle-${{ hashFiles('**/*.gradle*', '**/gradle-wrapper.properties') }}
-        restore-keys: ${{ runner.os }}-gradle-
-    - run: echo -e "$(ip addr show eth0 | grep "inet\b" | awk '{print $2}' | cut -d/ -f1)\t$(hostname -f) $(hostname -s)" | sudo tee -a /etc/hosts
-    - run: ./gradlew -DsparkVersions= -DhiveVersions=3 -DflinkVersions= -DkafkaVersions= -Pquick=true :iceberg-hive3-orc-bundle:check :iceberg-hive3:check :iceberg-hive-runtime:check -x javadoc
-    - uses: actions/upload-artifact@v4
-      if: failure()
-      with:
-        name: test logs
-        path: |
-          **/build/testlogs
+          **/build/testlogs

.github/workflows/kafka-connect-ci.yml

@@ -49,9 +49,6 @@ on:
     - 'dev/**'
     - 'mr/**'
     - 'flink/**'
-    - 'hive3/**'
-    - 'hive3-orc-bundle/**'
-    - 'hive-runtime/**'
     - 'spark/**'
     - 'docs/**'
     - 'site/**'

.github/workflows/spark-ci.yml

@@ -49,9 +49,6 @@ on:
     - 'dev/**'
     - 'site/**'
     - 'mr/**'
-    - 'hive3/**'
-    - 'hive3-orc-bundle/**'
-    - 'hive-runtime/**'
     - 'flink/**'
     - 'kafka-connect/**'
     - 'docs/**'

LICENSE

@@ -272,7 +272,7 @@ License: https://www.apache.org/licenses/LICENSE-2.0
 
 This product includes code from Apache Hive.
 
-* Hive metastore derby schema in hive-schema-3.1.0.derby.sql
+* Hive metastore derby schema in hive-schema-4.0.0.derby.sql
 
 Copyright: 2011-2018 The Apache Software Foundation
 Home page: https://hive.apache.org/

build.gradle

@@ -593,6 +593,8 @@ project(':iceberg-delta-lake') {
       }
       integrationImplementation project(path: ':iceberg-hive-metastore')
       integrationImplementation project(path: ':iceberg-hive-metastore', configuration: 'testArtifacts')
+      integrationImplementation(libs.hive4.metastore)
+      integrationImplementation(libs.hive4.standalone.metastore.server)
       integrationImplementation("org.apache.spark:spark-hive_${scalaVersion}:${libs.versions.spark.hive35.get()}") {
         exclude group: 'org.apache.avro', module: 'avro'
         exclude group: 'org.apache.arrow'
@@ -601,6 +603,7 @@ project(':iceberg-delta-lake') {
         exclude group: 'io.netty', module: 'netty-buffer'
         exclude group: 'io.netty', module: 'netty-common'
         exclude group: 'org.roaringbitmap'
+        exclude group: 'org.apache.hive'
       }
     }
   }
@@ -675,7 +678,7 @@ project(':iceberg-hive-metastore') {
 
     compileOnly libs.avro.avro
 
-    compileOnly(libs.hive2.metastore) {
+    compileOnly(libs.hive4.metastore) {
       exclude group: 'org.apache.avro', module: 'avro'
       exclude group: 'org.slf4j', module: 'slf4j-log4j12'
       exclude group: 'org.pentaho' // missing dependency
@@ -695,7 +698,7 @@ project(':iceberg-hive-metastore') {
     // that's really old. We use the core classifier to be able to override our guava
     // version. Luckily, hive-exec seems to work okay so far with this version of guava
     // See: https://github.com/apache/hive/blob/master/ql/pom.xml#L911 for more context.
-    testImplementation("${libs.hive2.exec.get().module}:${libs.hive2.exec.get().getVersion()}:core") {
+    testImplementation("${libs.hive4.exec.get().module}:${libs.hive4.exec.get().getVersion()}:core") {
       exclude group: 'org.apache.avro', module: 'avro'
       exclude group: 'org.slf4j', module: 'slf4j-log4j12'
       exclude group: 'org.pentaho' // missing dependency
@@ -707,7 +710,7 @@ project(':iceberg-hive-metastore') {
       exclude group: 'com.google.code.findbugs', module: 'jsr305'
     }
 
-    testImplementation(libs.hive2.metastore) {
+    testImplementation(libs.hive4.metastore) {
       exclude group: 'org.apache.avro', module: 'avro'
       exclude group: 'org.slf4j', module: 'slf4j-log4j12'
       exclude group: 'org.pentaho' // missing dependency
@@ -723,7 +726,9 @@ project(':iceberg-hive-metastore') {
       exclude group: 'com.zaxxer', module: 'HikariCP'
     }
 
-    compileOnly(libs.hadoop2.client) {
+    testImplementation(libs.hive4.standalone.metastore.server)
+
+    compileOnly(libs.hadoop3.client) {
       exclude group: 'org.apache.avro', module: 'avro'
       exclude group: 'org.slf4j', module: 'slf4j-log4j12'
     }

delta-lake/src/integration/java/org/apache/iceberg/delta/SparkDeltaLakeSnapshotTestBase.java

@@ -43,8 +43,8 @@ public static void startMetastoreAndSpark() {
             .master("local[2]")
             .config(SQLConf.PARTITION_OVERWRITE_MODE().key(), "dynamic")
             .config(
-                "spark.hadoop." + HiveConf.ConfVars.METASTOREURIS.varname,
-                hiveConf.get(HiveConf.ConfVars.METASTOREURIS.varname))
+                "spark.hadoop." + HiveConf.ConfVars.METASTORE_URIS.varname,
+                hiveConf.get(HiveConf.ConfVars.METASTORE_URIS.varname))
             .config("spark.sql.legacy.respectNullabilityInTextDatasetConversion", "true")
             .config("spark.databricks.delta.retentionDurationCheck.enabled", "false")
             .config("spark.sql.extensions", "io.delta.sql.DeltaSparkSessionExtension")

flink/v1.18/build.gradle

@@ -88,7 +88,7 @@ project(":iceberg-flink:iceberg-flink-${flinkMajorVersion}") {
     // that's really old. We use the core classifier to be able to override our guava
     // version. Luckily, hive-exec seems to work okay so far with this version of guava
     // See: https://github.com/apache/hive/blob/master/ql/pom.xml#L911 for more context.
-    testImplementation("${libs.hive2.exec.get().module}:${libs.hive2.exec.get().getVersion()}:core") {
+    testImplementation("${libs.hive4.exec.get().module}:${libs.hive4.exec.get().getVersion()}:core") {
       exclude group: 'org.apache.avro', module: 'avro'
       exclude group: 'org.slf4j', module: 'slf4j-log4j12'
       exclude group: 'org.pentaho' // missing dependency
@@ -100,7 +100,7 @@ project(":iceberg-flink:iceberg-flink-${flinkMajorVersion}") {
       exclude group: 'com.google.code.findbugs', module: 'jsr305'
     }
 
-    testImplementation(libs.hive2.metastore) {
+    testImplementation(libs.hive4.metastore) {
       exclude group: 'org.apache.avro', module: 'avro'
       exclude group: 'org.slf4j', module: 'slf4j-log4j12'
       exclude group: 'org.pentaho' // missing dependency
@@ -192,7 +192,7 @@ project(":iceberg-flink:iceberg-flink-runtime-${flinkMajorVersion}") {
       exclude group: 'org.apache.avro', module: 'avro'
     }
 
-    integrationImplementation(libs.hive2.metastore) {
+    integrationImplementation(libs.hive4.metastore) {
       exclude group: 'org.apache.avro', module: 'avro'
       exclude group: 'org.slf4j', module: 'slf4j-log4j12'
       exclude group: 'org.pentaho' // missing dependency
@@ -208,8 +208,9 @@ project(":iceberg-flink:iceberg-flink-runtime-${flinkMajorVersion}") {
       exclude group: 'com.zaxxer', module: 'HikariCP'
       exclude group: 'org.slf4j'
     }
+    integrationImplementation(libs.hive4.standalone.metastore.server)
 
-    integrationImplementation("${libs.hive2.exec.get().module}:${libs.hive2.exec.get().getVersion()}:core") {
+    integrationImplementation("${libs.hive4.exec.get().module}:${libs.hive4.exec.get().getVersion()}:core") {
       exclude group: 'org.apache.avro', module: 'avro'
       exclude group: 'org.slf4j', module: 'slf4j-log4j12'
       exclude group: 'org.pentaho' // missing dependency

flink/v1.18/flink/src/test/java/org/apache/iceberg/flink/CatalogTestBase.java

@@ -116,6 +116,6 @@ protected String getFullQualifiedTableName(String tableName) {
   }
 
   static String getURI(HiveConf conf) {
-    return conf.get(HiveConf.ConfVars.METASTOREURIS.varname);
+    return conf.get(HiveConf.ConfVars.METASTORE_URIS.varname);
   }
 }

flink/v1.18/flink/src/test/java/org/apache/iceberg/flink/TestFlinkHiveCatalog.java

@@ -57,7 +57,8 @@ public void testCreateCatalogWithHiveConfDir() throws IOException {
       Configuration newConf = new Configuration(hiveConf);
       // Set another new directory which is different with the hive metastore's warehouse path.
       newConf.set(
-          HiveConf.ConfVars.METASTOREWAREHOUSE.varname, "file://" + warehouseDir.getAbsolutePath());
+          HiveConf.ConfVars.METASTORE_WAREHOUSE.varname,
+          "file://" + warehouseDir.getAbsolutePath());
       newConf.writeXml(fos);
     }
     assertThat(hiveSiteXML.toPath()).exists();

flink/v1.18/flink/src/test/java/org/apache/iceberg/flink/source/TestFlinkInputFormatReaderDeletes.java

@@ -44,8 +44,8 @@ protected StructLikeSet rowSet(String tableName, Table testTable, String... colu
     Map<String, String> properties = Maps.newHashMap();
     properties.put(
         CatalogProperties.WAREHOUSE_LOCATION,
-        hiveConf.get(HiveConf.ConfVars.METASTOREWAREHOUSE.varname));
-    properties.put(CatalogProperties.URI, hiveConf.get(HiveConf.ConfVars.METASTOREURIS.varname));
+        hiveConf.get(HiveConf.ConfVars.METASTORE_WAREHOUSE.varname));
+    properties.put(CatalogProperties.URI, hiveConf.get(HiveConf.ConfVars.METASTORE_URIS.varname));
     properties.put(
         CatalogProperties.CLIENT_POOL_SIZE,
         Integer.toString(hiveConf.getInt("iceberg.hive.client-pool-size", 5)));

flink/v1.18/flink/src/test/java/org/apache/iceberg/flink/source/TestIcebergSourceReaderDeletes.java

@@ -62,8 +62,8 @@ protected StructLikeSet rowSet(String tableName, Table testTable, String... colu
     Map<String, String> properties = Maps.newHashMap();
     properties.put(
         CatalogProperties.WAREHOUSE_LOCATION,
-        hiveConf.get(HiveConf.ConfVars.METASTOREWAREHOUSE.varname));
-    properties.put(CatalogProperties.URI, hiveConf.get(HiveConf.ConfVars.METASTOREURIS.varname));
+        hiveConf.get(HiveConf.ConfVars.METASTORE_WAREHOUSE.varname));
+    properties.put(CatalogProperties.URI, hiveConf.get(HiveConf.ConfVars.METASTORE_URIS.varname));
     properties.put(
         CatalogProperties.CLIENT_POOL_SIZE,
         Integer.toString(hiveConf.getInt("iceberg.hive.client-pool-size", 5)));

flink/v1.19/build.gradle

@@ -88,7 +88,7 @@ project(":iceberg-flink:iceberg-flink-${flinkMajorVersion}") {
     // that's really old. We use the core classifier to be able to override our guava
     // version. Luckily, hive-exec seems to work okay so far with this version of guava
     // See: https://github.com/apache/hive/blob/master/ql/pom.xml#L911 for more context.
-    testImplementation("${libs.hive2.exec.get().module}:${libs.hive2.exec.get().getVersion()}:core") {
+    testImplementation("${libs.hive4.exec.get().module}:${libs.hive4.exec.get().getVersion()}:core") {
       exclude group: 'org.apache.avro', module: 'avro'
       exclude group: 'org.slf4j', module: 'slf4j-log4j12'
       exclude group: 'org.pentaho' // missing dependency
@@ -100,7 +100,7 @@ project(":iceberg-flink:iceberg-flink-${flinkMajorVersion}") {
       exclude group: 'com.google.code.findbugs', module: 'jsr305'
     }
 
-    testImplementation(libs.hive2.metastore) {
+    testImplementation(libs.hive4.metastore) {
       exclude group: 'org.apache.avro', module: 'avro'
       exclude group: 'org.slf4j', module: 'slf4j-log4j12'
       exclude group: 'org.pentaho' // missing dependency
@@ -116,6 +116,7 @@ project(":iceberg-flink:iceberg-flink-${flinkMajorVersion}") {
       exclude group: 'com.zaxxer', module: 'HikariCP'
       exclude group: 'org.slf4j'
     }
+    testImplementation(libs.hive4.standalone.metastore.server)
 
     testImplementation libs.awaitility
     testImplementation libs.assertj.core
@@ -193,7 +194,7 @@ project(":iceberg-flink:iceberg-flink-runtime-${flinkMajorVersion}") {
       exclude group: 'org.apache.avro', module: 'avro'
     }
 
-    integrationImplementation(libs.hive2.metastore) {
+    integrationImplementation(libs.hive4.metastore) {
       exclude group: 'org.apache.avro', module: 'avro'
       exclude group: 'org.slf4j', module: 'slf4j-log4j12'
       exclude group: 'org.pentaho' // missing dependency
@@ -210,7 +211,7 @@ project(":iceberg-flink:iceberg-flink-runtime-${flinkMajorVersion}") {
       exclude group: 'org.slf4j'
     }
 
-    integrationImplementation("${libs.hive2.exec.get().module}:${libs.hive2.exec.get().getVersion()}:core") {
+    integrationImplementation("${libs.hive4.exec.get().module}:${libs.hive4.exec.get().getVersion()}:core") {
       exclude group: 'org.apache.avro', module: 'avro'
       exclude group: 'org.slf4j', module: 'slf4j-log4j12'
       exclude group: 'org.pentaho' // missing dependency

flink/v1.19/flink/src/test/java/org/apache/iceberg/flink/CatalogTestBase.java

@@ -116,6 +116,6 @@ protected String getFullQualifiedTableName(String tableName) {
   }
 
   static String getURI(HiveConf conf) {
-    return conf.get(HiveConf.ConfVars.METASTOREURIS.varname);
+    return conf.get(HiveConf.ConfVars.METASTORE_URIS.varname);
   }
 }

flink/v1.19/flink/src/test/java/org/apache/iceberg/flink/TestFlinkHiveCatalog.java

@@ -57,7 +57,8 @@ public void testCreateCatalogWithHiveConfDir() throws IOException {
       Configuration newConf = new Configuration(hiveConf);
       // Set another new directory which is different with the hive metastore's warehouse path.
       newConf.set(
-          HiveConf.ConfVars.METASTOREWAREHOUSE.varname, "file://" + warehouseDir.getAbsolutePath());
+          HiveConf.ConfVars.METASTORE_WAREHOUSE.varname,
+          "file://" + warehouseDir.getAbsolutePath());
       newConf.writeXml(fos);
     }
     assertThat(hiveSiteXML.toPath()).exists();

flink/v1.19/flink/src/test/java/org/apache/iceberg/flink/source/TestFlinkInputFormatReaderDeletes.java

@@ -44,8 +44,8 @@ protected StructLikeSet rowSet(String tableName, Table testTable, String... colu
     Map<String, String> properties = Maps.newHashMap();
     properties.put(
         CatalogProperties.WAREHOUSE_LOCATION,
-        hiveConf.get(HiveConf.ConfVars.METASTOREWAREHOUSE.varname));
-    properties.put(CatalogProperties.URI, hiveConf.get(HiveConf.ConfVars.METASTOREURIS.varname));
+        hiveConf.get(HiveConf.ConfVars.METASTORE_WAREHOUSE.varname));
+    properties.put(CatalogProperties.URI, hiveConf.get(HiveConf.ConfVars.METASTORE_URIS.varname));
     properties.put(
         CatalogProperties.CLIENT_POOL_SIZE,
         Integer.toString(hiveConf.getInt("iceberg.hive.client-pool-size", 5)));

flink/v1.19/flink/src/test/java/org/apache/iceberg/flink/source/TestIcebergSourceReaderDeletes.java

@@ -62,8 +62,8 @@ protected StructLikeSet rowSet(String tableName, Table testTable, String... colu
     Map<String, String> properties = Maps.newHashMap();
     properties.put(
         CatalogProperties.WAREHOUSE_LOCATION,
-        hiveConf.get(HiveConf.ConfVars.METASTOREWAREHOUSE.varname));
-    properties.put(CatalogProperties.URI, hiveConf.get(HiveConf.ConfVars.METASTOREURIS.varname));
+        hiveConf.get(HiveConf.ConfVars.METASTORE_WAREHOUSE.varname));
+    properties.put(CatalogProperties.URI, hiveConf.get(HiveConf.ConfVars.METASTORE_URIS.varname));
     properties.put(
         CatalogProperties.CLIENT_POOL_SIZE,
         Integer.toString(hiveConf.getInt("iceberg.hive.client-pool-size", 5)));

flink/v1.20/build.gradle

@@ -88,7 +88,7 @@ project(":iceberg-flink:iceberg-flink-${flinkMajorVersion}") {
     // that's really old. We use the core classifier to be able to override our guava
     // version. Luckily, hive-exec seems to work okay so far with this version of guava
     // See: https://github.com/apache/hive/blob/master/ql/pom.xml#L911 for more context.
-    testImplementation("${libs.hive2.exec.get().module}:${libs.hive2.exec.get().getVersion()}:core") {
+    testImplementation("${libs.hive4.exec.get().module}:${libs.hive4.exec.get().getVersion()}:core") {
       exclude group: 'org.apache.avro', module: 'avro'
       exclude group: 'org.slf4j', module: 'slf4j-log4j12'
       exclude group: 'org.pentaho' // missing dependency
@@ -100,7 +100,7 @@ project(":iceberg-flink:iceberg-flink-${flinkMajorVersion}") {
       exclude group: 'com.google.code.findbugs', module: 'jsr305'
     }
 
-    testImplementation(libs.hive2.metastore) {
+    testImplementation(libs.hive4.metastore) {
       exclude group: 'org.apache.avro', module: 'avro'
       exclude group: 'org.slf4j', module: 'slf4j-log4j12'
       exclude group: 'org.pentaho' // missing dependency
@@ -116,6 +116,7 @@ project(":iceberg-flink:iceberg-flink-${flinkMajorVersion}") {
       exclude group: 'com.zaxxer', module: 'HikariCP'
       exclude group: 'org.slf4j'
     }
+    testImplementation(libs.hive4.standalone.metastore.server)
 
     testImplementation libs.awaitility
     testImplementation libs.assertj.core
@@ -193,7 +194,7 @@ project(":iceberg-flink:iceberg-flink-runtime-${flinkMajorVersion}") {
       exclude group: 'org.apache.avro', module: 'avro'
     }
 
-    integrationImplementation(libs.hive2.metastore) {
+    integrationImplementation(libs.hive4.metastore) {
       exclude group: 'org.apache.avro', module: 'avro'
       exclude group: 'org.slf4j', module: 'slf4j-log4j12'
       exclude group: 'org.pentaho' // missing dependency
@@ -210,7 +211,7 @@ project(":iceberg-flink:iceberg-flink-runtime-${flinkMajorVersion}") {
       exclude group: 'org.slf4j'
     }
 
-    integrationImplementation("${libs.hive2.exec.get().module}:${libs.hive2.exec.get().getVersion()}:core") {
+    integrationImplementation("${libs.hive4.exec.get().module}:${libs.hive4.exec.get().getVersion()}:core") {
       exclude group: 'org.apache.avro', module: 'avro'
       exclude group: 'org.slf4j', module: 'slf4j-log4j12'
       exclude group: 'org.pentaho' // missing dependency

flink/v1.20/flink/src/test/java/org/apache/iceberg/flink/CatalogTestBase.java

@@ -116,6 +116,6 @@ protected String getFullQualifiedTableName(String tableName) {
   }
 
   static String getURI(HiveConf conf) {
-    return conf.get(HiveConf.ConfVars.METASTOREURIS.varname);
+    return conf.get(HiveConf.ConfVars.METASTORE_URIS.varname);
   }
 }

flink/v1.20/flink/src/test/java/org/apache/iceberg/flink/TestFlinkHiveCatalog.java

@@ -57,7 +57,8 @@ public void testCreateCatalogWithHiveConfDir() throws IOException {
       Configuration newConf = new Configuration(hiveConf);
       // Set another new directory which is different with the hive metastore's warehouse path.
       newConf.set(
-          HiveConf.ConfVars.METASTOREWAREHOUSE.varname, "file://" + warehouseDir.getAbsolutePath());
+          HiveConf.ConfVars.METASTORE_WAREHOUSE.varname,
+          "file://" + warehouseDir.getAbsolutePath());
       newConf.writeXml(fos);
     }
     assertThat(hiveSiteXML.toPath()).exists();