diff --git a/.github/actions/java-test/action.yaml b/.github/actions/java-test/action.yaml
index 1f7899f8ac..24af6908ce 100644
--- a/.github/actions/java-test/action.yaml
+++ b/.github/actions/java-test/action.yaml
@@ -29,6 +29,10 @@ inputs:
     description: 'Maven options passed to the mvn command'
     required: false
     default: ''
+  features:
+    description: 'Build native features (like hdfs e.g)'
+    required: false
+    default: ''
   scan_impl:
     description: 'The default Parquet scan implementation'
     required: false
@@ -45,7 +49,11 @@ runs:
       shell: bash
       run: |
         cd native
-        cargo build
+        FEATURES_ARG=""
+        if [ -n "${{ inputs.features }}" ]; then
+          FEATURES_ARG="--features=${{ inputs.features  }}"
+        fi
+        cargo build $FEATURES_ARG
 
     - name: Cache Maven dependencies
       uses: actions/cache@v4
@@ -82,6 +90,10 @@ runs:
         MAVEN_SUITES="$(echo "${{ inputs.suites }}" | paste -sd, -)"
         echo "Running with MAVEN_SUITES=$MAVEN_SUITES"
         MAVEN_OPTS="-Xmx4G -Xms2G -DwildcardSuites=$MAVEN_SUITES -XX:+UnlockDiagnosticVMOptions -XX:+ShowMessageBoxOnError -XX:+HeapDumpOnOutOfMemoryError -XX:ErrorFile=./hs_err_pid%p.log" SPARK_HOME=`pwd` ./mvnw -B clean install ${{ inputs.maven_opts }}
+        # Hadoop mini cluster may interfere with spark shutdown hook, causing the tmp directory to not be fully cleaned up.
+        # Clean tmp dir to avoid hashFiles error
+        ls spark/target/tmp
+        rm -rf spark/target/tmp
     - name: Upload crash logs
       if: failure()
       uses: actions/upload-artifact@v4
diff --git a/.github/workflows/pr_build_linux.yml b/.github/workflows/pr_build_linux.yml
index c45355978e..c5651dc101 100644
--- a/.github/workflows/pr_build_linux.yml
+++ b/.github/workflows/pr_build_linux.yml
@@ -170,4 +170,42 @@ jobs:
           suites: ${{ matrix.suite.value }}
           maven_opts: ${{ matrix.profile.maven_opts }}
           scan_impl: ${{ matrix.profile.scan_impl }}
-          upload-test-reports: true
\ No newline at end of file
+          upload-test-reports: true
+
+  # Java tests with native features
+  linux-test-features:
+    strategy:
+      matrix:
+        os: [ubuntu-latest]
+        java_version: [17]
+        features:
+          - value: "hdfs-opendal"
+            suites: |
+              org.apache.comet.parquet.ParquetReadFromFakeHadoopFsSuite
+              org.apache.comet.parquet.ParquetReadFromHdfsSuite
+          - value: "hdfs"
+            suites: |
+              org.apache.comet.parquet.ParquetReadFromHdfsSuite
+      fail-fast: false
+    name: ${{ matrix.os }}/java ${{ matrix.java_version }}-features [${{ matrix.features.value }}]
+    runs-on: ${{ matrix.os }}
+    container:
+      image: amd64/rust
+      env:
+        JAVA_TOOL_OPTIONS: ${{ matrix.profile.java_version == '17' && '--add-exports=java.base/sun.nio.ch=ALL-UNNAMED --add-exports=java.base/sun.util.calendar=ALL-UNNAMED --add-opens=java.base/java.nio=ALL-UNNAMED --add-opens=java.base/java.lang=ALL-UNNAMED' || '' }}
+
+    steps:
+      - uses: actions/checkout@v5
+      - name: Setup Rust & Java toolchain
+        uses: ./.github/actions/setup-builder
+        with:
+          rust-version: ${{env.RUST_VERSION}}
+          jdk-version: ${{ matrix.java_version }}
+      - name: Java test steps
+        uses: ./.github/actions/java-test
+        with:
+          artifact_name: ${{ matrix.os }}-java-${{ matrix.java_version }}-features-${{ matrix.features.value }}-${{ github.run_id }}-${{ github.run_number }}-${{ github.run_attempt }}
+          features: ${{ matrix.features.value }}
+          maven_opts: "-Dtest=none"
+          suites: ${{ matrix.features.suites }}
+          upload-test-reports: true
diff --git a/.github/workflows/pr_build_macos.yml b/.github/workflows/pr_build_macos.yml
index 6c71006e5f..7dab979b36 100644
--- a/.github/workflows/pr_build_macos.yml
+++ b/.github/workflows/pr_build_macos.yml
@@ -131,3 +131,39 @@ jobs:
           artifact_name: ${{ matrix.os }}-${{ matrix.profile.name }}-${{ matrix.suite.name }}-${{ github.run_id }}-${{ github.run_number }}-${{ github.run_attempt }}
           suites: ${{ matrix.suite.value }}
           maven_opts: ${{ matrix.profile.maven_opts }}
+
+  # Java tests with native features
+  macos-aarch64-test-features:
+    strategy:
+      matrix:
+        os: [macos-14]
+        java_version: [17]
+        features:
+          - value: "hdfs-opendal"
+            suites: |
+              org.apache.comet.parquet.ParquetReadFromFakeHadoopFsSuite
+              org.apache.comet.parquet.ParquetReadFromHdfsSuite
+          - value: "hdfs"
+            suites: |
+              org.apache.comet.parquet.ParquetReadFromHdfsSuite
+      fail-fast: false
+    name: ${{ matrix.os }}/java ${{ matrix.java_version }}-features [${{ matrix.features.value }}]
+    runs-on: ${{ matrix.os }}
+
+    steps:
+      - uses: actions/checkout@v5
+      - name: Setup Rust & Java toolchain
+        uses: ./.github/actions/setup-macos-builder
+        with:
+          rust-version: ${{env.RUST_VERSION}}
+          jdk-version: ${{ matrix.java_version }}
+          jdk-architecture: aarch64
+          protoc-architecture: aarch_64
+      - name: Java test steps
+        uses: ./.github/actions/java-test
+        with:
+          artifact_name: ${{ matrix.os }}-java-${{ matrix.java_version }}-features-${{ matrix.features.value }}-${{ github.run_id }}-${{ github.run_number }}-${{ github.run_attempt }}
+          features: ${{ matrix.features.value }}
+          maven_opts: "-Dtest=none"
+          suites: ${{ matrix.features.suites }}
+          upload-test-reports: true
diff --git a/native/core/Cargo.toml b/native/core/Cargo.toml
index 16a8a7316d..e6828237ae 100644
--- a/native/core/Cargo.toml
+++ b/native/core/Cargo.toml
@@ -74,8 +74,8 @@ aws-credential-types = { workspace = true }
 parking_lot = "0.12.3"
 datafusion-comet-objectstore-hdfs = { path = "../hdfs", optional = true, default-features = false, features = ["hdfs"] }
 
-object_store_opendal = {version = "0.54.0", optional = true}
-hdfs-sys = {version = "0.3", optional = true, features = ["hdfs_3_3"]}
+object_store_opendal = { version = "0.54.0", optional = true }
+hdfs-sys = { version = "0.3", optional = true, features = ["hdfs_3_3"] }
 opendal = { version ="0.54.0", optional = true, features = ["services-hdfs"] }
 
 [target.'cfg(target_os = "linux")'.dependencies]
diff --git a/spark/src/test/scala/org/apache/comet/WithHdfsCluster.scala b/spark/src/test/scala/org/apache/comet/WithHdfsCluster.scala
index 49124d63e5..30d1581696 100644
--- a/spark/src/test/scala/org/apache/comet/WithHdfsCluster.scala
+++ b/spark/src/test/scala/org/apache/comet/WithHdfsCluster.scala
@@ -74,6 +74,9 @@ trait WithHdfsCluster extends Logging {
   }
 
   def stopHdfsCluster(): Unit = {
+    // wait for hdfs async blocking thread exit to avoid jvm crash, see:
+    //  https://github.com/apache/datafusion-comet/issues/2354
+    Thread.sleep(2000)
     if (hdfsCluster != null) hdfsCluster.shutdown(true)
     if (hadoopConfDir != null) FileUtils.deleteDirectory(hadoopConfDir)
   }
diff --git a/spark/src/test/scala/org/apache/comet/parquet/ParquetReadFromFakeHadoopFsSuite.scala b/spark/src/test/scala/org/apache/comet/parquet/ParquetReadFromFakeHadoopFsSuite.scala
index f4a8b5ed82..03f9437b29 100644
--- a/spark/src/test/scala/org/apache/comet/parquet/ParquetReadFromFakeHadoopFsSuite.scala
+++ b/spark/src/test/scala/org/apache/comet/parquet/ParquetReadFromFakeHadoopFsSuite.scala
@@ -26,7 +26,7 @@ import java.util.UUID
 import org.apache.commons.io.FileUtils
 import org.apache.spark.SparkConf
 import org.apache.spark.sql.{CometTestBase, DataFrame, SaveMode}
-import org.apache.spark.sql.comet.CometNativeScanExec
+import org.apache.spark.sql.comet.{CometNativeScanExec, CometScanExec}
 import org.apache.spark.sql.execution.adaptive.AdaptiveSparkPlanHelper
 import org.apache.spark.sql.functions.{col, sum}
 
@@ -45,10 +45,10 @@ class ParquetReadFromFakeHadoopFsSuite extends CometTestBase with AdaptiveSparkP
   }
 
   override def beforeAll(): Unit = {
-    // Initialize fake root dir
-    fake_root_dir = Files.createTempDirectory(s"comet_fake_${UUID.randomUUID().toString}").toFile
     // Initialize Spark session
     super.beforeAll()
+    // Initialize fake root dir
+    fake_root_dir = Files.createTempDirectory(s"comet_fake_${UUID.randomUUID().toString}").toFile
   }
 
   protected override def afterAll(): Unit = {
@@ -62,16 +62,21 @@ class ParquetReadFromFakeHadoopFsSuite extends CometTestBase with AdaptiveSparkP
   }
 
   private def assertCometNativeScanOnFakeFs(df: DataFrame): Unit = {
-    val scans = collect(df.queryExecution.executedPlan) { case p: CometNativeScanExec =>
-      p
+    val scans = collect(df.queryExecution.executedPlan) {
+      case p: CometNativeScanExec =>
+        assert(
+          p.nativeOp.getNativeScan
+            .getFilePartitions(0)
+            .getPartitionedFile(0)
+            .getFilePath
+            .startsWith(FakeHDFSFileSystem.PREFIX))
+        p
+      case p: CometScanExec if p.scanImpl == CometConf.SCAN_NATIVE_ICEBERG_COMPAT =>
+        assert(p.toString().contains(FakeHDFSFileSystem.PREFIX))
+        p
     }
     assert(scans.size == 1)
-    assert(
-      scans.head.nativeOp.getNativeScan
-        .getFilePartitions(0)
-        .getPartitionedFile(0)
-        .getFilePath
-        .startsWith(FakeHDFSFileSystem.PREFIX))
+
   }
 
   // This test fails for 'hdfs' but succeeds for 'open-dal'. 'hdfs' requires this fix
@@ -82,10 +87,13 @@ class ParquetReadFromFakeHadoopFsSuite extends CometTestBase with AdaptiveSparkP
     val testFilePath =
       s"${FakeHDFSFileSystem.PREFIX}${fake_root_dir.getAbsolutePath}/data/test-file.parquet"
     writeTestParquetFile(testFilePath)
-    withSQLConf(CometConf.COMET_NATIVE_SCAN_IMPL.key -> CometConf.SCAN_NATIVE_DATAFUSION) {
-      val df = spark.read.format("parquet").load(testFilePath).agg(sum(col("id")))
-      assertCometNativeScanOnFakeFs(df)
-      assert(df.first().getLong(0) == 499500)
+    Seq(CometConf.SCAN_NATIVE_DATAFUSION, CometConf.SCAN_NATIVE_ICEBERG_COMPAT).foreach {
+      scanImpl =>
+        withSQLConf(CometConf.COMET_NATIVE_SCAN_IMPL.key -> scanImpl) {
+          val df = spark.read.format("parquet").load(testFilePath).agg(sum(col("id")))
+          assertCometNativeScanOnFakeFs(df)
+          assert(df.first().getLong(0) == 499500)
+        }
     }
   }
 }
diff --git a/spark/src/test/scala/org/apache/comet/parquet/ParquetReadFromHdfsSuite.scala b/spark/src/test/scala/org/apache/comet/parquet/ParquetReadFromHdfsSuite.scala
new file mode 100644
index 0000000000..719e03d6e9
--- /dev/null
+++ b/spark/src/test/scala/org/apache/comet/parquet/ParquetReadFromHdfsSuite.scala
@@ -0,0 +1,87 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements.  See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership.  The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License.  You may obtain a copy of the License at
+ *
+ *   http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing,
+ * software distributed under the License is distributed on an
+ * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
+ * KIND, either express or implied.  See the License for the
+ * specific language governing permissions and limitations
+ * under the License.
+ */
+
+package org.apache.comet.parquet
+
+import org.apache.spark.sql.{CometTestBase, DataFrame, SaveMode}
+import org.apache.spark.sql.comet.{CometNativeScanExec, CometScanExec}
+import org.apache.spark.sql.execution.adaptive.AdaptiveSparkPlanHelper
+import org.apache.spark.sql.functions.{col, sum}
+
+import org.apache.comet.{CometConf, WithHdfsCluster}
+
+class ParquetReadFromHdfsSuite
+    extends CometTestBase
+    with AdaptiveSparkPlanHelper
+    with WithHdfsCluster {
+
+  override protected def createSparkSession: SparkSessionType = {
+    val sparkSession = super.createSparkSession
+    // start HDFS cluster and add hadoop conf
+    startHdfsCluster()
+    sparkSession.sparkContext.hadoopConfiguration.addResource(getHadoopConfFile)
+    sparkSession
+  }
+
+  protected override def afterAll(): Unit = {
+    super.afterAll()
+    stopHdfsCluster()
+  }
+
+  private def writeTestParquetFile(filePath: String): Unit = {
+    val df = spark.range(0, 1000)
+    df.write.format("parquet").mode(SaveMode.Overwrite).save(filePath)
+  }
+
+  private def assertCometNativeScanOnHDFS(df: DataFrame): Unit = {
+    val scans = collect(df.queryExecution.executedPlan) {
+      case p: CometNativeScanExec =>
+        assert(
+          p.nativeOp.getNativeScan
+            .getFilePartitions(0)
+            .getPartitionedFile(0)
+            .getFilePath
+            .startsWith("hdfs://"))
+        p
+      case p: CometScanExec if p.scanImpl == CometConf.SCAN_NATIVE_ICEBERG_COMPAT =>
+        assert(p.toString().contains("hdfs://"))
+        p
+    }
+    assert(scans.size == 1)
+  }
+
+  test("test native_datafusion scan on hdfs") {
+    assume(isFeatureEnabled("hdfs") || isFeatureEnabled("hdfs-opendal"))
+
+    withTmpHdfsDir { dir =>
+      {
+        val testFilePath = dir.toString
+        writeTestParquetFile(testFilePath)
+        Seq(CometConf.SCAN_NATIVE_DATAFUSION, CometConf.SCAN_NATIVE_ICEBERG_COMPAT).foreach {
+          scanImpl =>
+            withSQLConf(CometConf.COMET_NATIVE_SCAN_IMPL.key -> scanImpl) {
+              val df = spark.read.format("parquet").load(testFilePath).agg(sum(col("id")))
+              assertCometNativeScanOnHDFS(df)
+              assert(df.first().getLong(0) == 499500)
+            }
+        }
+      }
+    }
+  }
+}