feat: Add support for native ice binary using graalvm

.bin/agent-ice

@@ -0,0 +1,17 @@
+#!/usr/bin/env bash
+set -euo pipefail
+SCRIPT_DIR=$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd)
+PROJECT_ROOT=$(cd -- "$SCRIPT_DIR/.." && pwd)
+
+shopt -s nullglob
+jars=("$PROJECT_ROOT"/ice/target/ice-*-shaded.jar)
+shopt -u nullglob
+if (( ${#jars[@]} == 0 )); then
+  echo "agent-ice: no ice-*-shaded.jar in $PROJECT_ROOT/ice/target/." >&2
+  exit 1
+fi
+jar=$(ls -t "${jars[@]}" | head -n1)
+
+exec "${JAVA_HOME:?JAVA_HOME not set — run inside direnv shell}/bin/java" \
+-agentlib:native-image-agent=config-merge-dir=$PROJECT_ROOT/ice/src/main/resources/META-INF/native-image/com.altinity/ice,experimental-class-loader-support \
+-jar "$jar" "$@"

.github/workflows/verify.yaml

@@ -17,7 +17,6 @@ jobs:
       - run: ./mvnw clean verify
       - name: Install
         run: ./mvnw install
-      # TODO: check native-image can build ice
       - name: Run Scenario-Based Integration Tests
         run: ../mvnw test -Dtest=ScenarioBasedIT
         working-directory: ice-rest-catalog
@@ -58,3 +57,33 @@ jobs:
           -Dit.test=DockerScenarioBasedIT,DockerLocalFileIOClickHouseIT
           -Ddocker.image=altinity/ice-rest-catalog:debug-with-ice-latest-master-amd64
           -Dclickhouse.image=altinity/clickhouse-server:25.8.16.20002.altinityantalya
+  native-build:
+    name: Build and test native image (${{ matrix.arch }})
+    runs-on: ${{ matrix.runner }}
+    strategy:
+      matrix:
+        include:
+          - arch: amd64
+            runner: ubuntu-24.04
+            artifact: ice-native-amd64-dynamic
+          - arch: arm64
+            runner: ubuntu-24.04-arm
+            artifact: ice-native-arm64-dynamic
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-java@v4
+        with:
+          java-version: '21'
+          distribution: 'graalvm'
+          cache: maven
+      - name: Build native image (dynamic linking - no musl required)
+        run: ./mvnw -Pnative clean install -Dmaven.test.skip=true
+      - name: Run Scenario-Based Integration Tests
+        run: ../mvnw test -Dtest=ScenarioBasedIT -Dice.native=true
+        working-directory: ice-rest-catalog
+      - name: Upload native binary as artifact
+        uses: actions/upload-artifact@v4
+        with:
+          name: ${{ matrix.artifact }}
+          path: ice/target/ice
+          if-no-files-found: error

README.md

@@ -10,15 +10,23 @@ Create/delete tables, insert data with `ice insert -p ns1.table1 file://example.
 
 ## Installation
 
-Pre-built binaries\* (+ links to Docker images for [ice](https://hub.docker.com/r/altinity/ice) and [ice-rest-catalog](https://hub.docker.com/r/altinity/ice-rest-catalog)) are available from [GitHub Releases](https://github.com/Altinity/ice/releases) page.
-> \* currently require `java` 21+ to run (available [here](https://adoptium.net/installation/)).  
+Pre-built binaries (+ links to Docker images for [ice](https://hub.docker.com/r/altinity/ice) and [ice-rest-catalog](https://hub.docker.com/r/altinity/ice-rest-catalog)) are available from [GitHub Releases](https://github.com/Altinity/ice/releases) page.
+
+**Two types of binaries are available:**
+
+1. **Java-based binaries** - Require Java 21+ to run (available [here](https://adoptium.net/installation/))
+2. **Native binaries** - Standalone executables with no Java dependency
+   - `ice-native-amd64` - Dynamic binary for x86_64 Linux (requires glibc)
+   - `ice-native-arm64` - Dynamic binary for ARM64 Linux (requires glibc)
 
 ## Usage
 
 See [examples/](examples/).
 
 ## Development
 
+### Standard Build (Java-based)
+
 Install [sdkman](https://sdkman.io/install), then
 
 ```shell
@@ -35,6 +43,23 @@ sdk env
 ./mvnw
 ```
 
+### Native Image Build (Standalone)
+
+Build standalone native binaries with no Java dependency:
+
+```shell
+# Install prerequisites
+sdk env  # or ensure Java 21+ and GraalVM are available
+
+# Build ice binary (both amd64 and arm64)
+mvn -Pnative -pl ice clean package -Dmaven.test.skip=true
+
+# Docker builds
+docker build -f ice/Dockerfile.native -t ice-native:amd64 .
+
+docker build -f ice/Dockerfile.native -t ice-native:arm64 .
+```
+
 ## License
 
 Copyright (c) 2025, Altinity Inc and/or its affiliates. All rights reserved.  

examples/scratch/README.md

@@ -37,7 +37,7 @@ ice insert nyc.taxis_p_by_day -p \
 
 # delete partition
 ice delete nyc.taxis_p_by_day  \
-  --partition '[{"name": "tpep_pickup_datetime", "values": ["2024-12-31T23:51:20"]}]' --dry-run=false
+  --partition '[{"name": "tpep_pickup_datetime_day", "values": ["2024-12-31"]}]' --dry-run=false
 
 # insert data ordered by tpep_pickup_datetime column
 ice insert nyc.taxis_s_by_day -p \

ice-rest-catalog/pom.xml

@@ -17,6 +17,7 @@
     <jetcd.version>0.8.5</jetcd.version>
     <grpc.version>1.70.0</grpc.version>
     <testcontainers.version>1.21.4</testcontainers.version>
+    <ice.native>false</ice.native>
   </properties>
 
   <dependencies>
@@ -562,6 +563,9 @@
         <groupId>org.apache.maven.plugins</groupId>
         <artifactId>maven-failsafe-plugin</artifactId>
         <configuration>
+          <systemPropertyVariables>
+            <ice.native>${ice.native}</ice.native>
+          </systemPropertyVariables>
           <excludes>
             <exclude>**/DockerScenarioBasedIT.java</exclude>
             <exclude>**/DockerLocalFileIOClickHouseIT.java</exclude>

ice-rest-catalog/src/test/java/com/altinity/ice/rest/catalog/ScenarioBasedIT.java

@@ -23,6 +23,8 @@
  */
 public class ScenarioBasedIT extends RESTCatalogTestBase {
 
+  private static final boolean USE_NATIVE = Boolean.getBoolean("ice.native");
+
   @Override
   protected ScenarioTestRunner createScenarioRunner(String scenarioName) throws Exception {
     Path scenariosDir = getScenariosDirectory();
@@ -42,18 +44,24 @@ protected ScenarioTestRunner createScenarioRunner(String scenarioName) throws Ex
 
     // Try to find ice-jar in the build
     String projectRoot = Paths.get("").toAbsolutePath().getParent().toString();
-    String iceJar = projectRoot + "/ice/target/ice-jar";
+    String iceJar =
+        USE_NATIVE ? projectRoot + "/ice/target/ice" : projectRoot + "/ice/target/ice-jar";
     File iceJarFile = new File(iceJar);
 
-    if (iceJarFile.exists() && iceJarFile.canExecute()) {
-      // Use pre-built ice-jar if available
-      templateVars.put("ICE_CLI", iceJar);
-      logger.info("Using ice-jar from: {}", iceJar);
+    String override = System.getenv("ICE_CLI_OVERRIDE");
+    if (override != null) {
+      templateVars.put("ICE_CLI", override);
     } else {
-      // Fall back to using local-ice wrapper script
-      String localIce = projectRoot + "/.bin/local-ice";
-      templateVars.put("ICE_CLI", localIce);
-      logger.info("Using local-ice script from: {}", localIce);
+      if (iceJarFile.exists() && iceJarFile.canExecute()) {
+        // Use pre-built ice-jar if available
+        templateVars.put("ICE_CLI", iceJar);
+        logger.info("Using ice-jar from: {}", iceJar);
+      } else {
+        // Fall back to using local-ice wrapper script
+        String localIce = projectRoot + "/.bin/local-ice";
+        templateVars.put("ICE_CLI", localIce);
+        logger.info("Using local-ice script from: {}", localIce);
+      }
     }
 
     return new ScenarioTestRunner(scenariosDir, templateVars);

ice-rest-catalog/src/test/resources/scenarios/basic-operations/run.sh.tmpl

@@ -30,6 +30,10 @@ echo "OK list-namespaces listed ${NAMESPACE_NAME}"
 {{ICE_CLI}} --config {{CLI_CONFIG}} describe
 echo "OK Listed namespaces"
 
+# Check configuration
+{{ICE_CLI}} --config {{CLI_CONFIG}} check
+echo "OK checked configuration"
+
 # Insert from file (like README: ice insert flowers.iris -p file://...)
 {{ICE_CLI}} --config {{CLI_CONFIG}} insert ${TABLE_IRIS} -p "file://${INPUT_PATH}"
 echo "Inserted from file into ${TABLE_IRIS}"

ice-rest-catalog/src/test/resources/scenarios/describe-parquet/run.sh.tmpl

@@ -0,0 +1,151 @@
+#!/bin/bash
+set -e
+
+echo "Running describe-parquet test..."
+
+SCENARIO_DIR="{{SCENARIO_DIR}}"
+INPUT_PATH="${SCENARIO_DIR}/${INPUT_FILE}"
+TARGET="file://${INPUT_PATH}"
+
+# Default invocation: YAML, summary only
+DEFAULT_OUT=$({{ICE_CLI}} --config {{CLI_CONFIG}} describe-parquet "${TARGET}")
+for expected in "summary:" "rows: ${EXPECTED_ROWS}" "rowGroups:" "compressedSize:" "uncompressedSize:" "columnCount: ${EXPECTED_COLUMN_COUNT}" "createdBy:"; do
+  if ! grep -qF "${expected}" <<<"${DEFAULT_OUT}"; then
+    echo "FAIL default describe-parquet output missing '${expected}'"
+    printf '%s\n' "${DEFAULT_OUT}"
+    exit 1
+  fi
+done
+if grep -q "^columns:" <<<"${DEFAULT_OUT}"; then
+  echo "FAIL default describe-parquet output should not include columns section"
+  printf '%s\n' "${DEFAULT_OUT}"
+  exit 1
+fi
+echo "OK Default output verified"
+
+# -s / --summary
+SUMMARY_OUT=$({{ICE_CLI}} --config {{CLI_CONFIG}} describe-parquet -s "${TARGET}")
+if ! grep -qF "rows: ${EXPECTED_ROWS}" <<<"${SUMMARY_OUT}"; then
+  echo "FAIL --summary output missing expected row count"
+  printf '%s\n' "${SUMMARY_OUT}"
+  exit 1
+fi
+echo "OK --summary output verified"
+
+# --columns: each iris column with type and repetition
+COLUMNS_OUT=$({{ICE_CLI}} --config {{CLI_CONFIG}} describe-parquet --columns "${TARGET}")
+if ! grep -q "^columns:" <<<"${COLUMNS_OUT}"; then
+  echo "FAIL --columns output missing 'columns:' section"
+  printf '%s\n' "${COLUMNS_OUT}"
+  exit 1
+fi
+for col in "sepal.length" "sepal.width" "petal.length" "petal.width" "variety"; do
+  if ! grep -qF "name: ${col}" <<<"${COLUMNS_OUT}" && ! grep -qF "name: \"${col}\"" <<<"${COLUMNS_OUT}"; then
+    echo "FAIL --columns output missing column '${col}'"
+    printf '%s\n' "${COLUMNS_OUT}"
+    exit 1
+  fi
+done
+if ! grep -qE 'repetition: "?(REQUIRED|OPTIONAL|REPEATED)"?' <<<"${COLUMNS_OUT}"; then
+  echo "FAIL --columns output missing repetition field"
+  printf '%s\n' "${COLUMNS_OUT}"
+  exit 1
+fi
+if ! grep -qE 'type: "?(DOUBLE|FLOAT|INT32|INT64|BINARY|BOOLEAN|FIXED_LEN_BYTE_ARRAY|INT96)"?' <<<"${COLUMNS_OUT}"; then
+  echo "FAIL --columns output missing primitive type field"
+  printf '%s\n' "${COLUMNS_OUT}"
+  exit 1
+fi
+echo "OK --columns output verified"
+
+# -r / --row-groups (no per-column details)
+ROWGROUPS_OUT=$({{ICE_CLI}} --config {{CLI_CONFIG}} describe-parquet -r "${TARGET}")
+if ! grep -q "^rowGroups:" <<<"${ROWGROUPS_OUT}"; then
+  echo "FAIL -r output missing 'rowGroups:' section"
+  printf '%s\n' "${ROWGROUPS_OUT}"
+  exit 1
+fi
+for expected in "index:" "rowCount:" "totalSize:" "compressedSize:" "startingPos:"; do
+  if ! grep -qF "${expected}" <<<"${ROWGROUPS_OUT}"; then
+    echo "FAIL -r output missing '${expected}'"
+    printf '%s\n' "${ROWGROUPS_OUT}"
+    exit 1
+  fi
+done
+if grep -qE "encodings:|codec:" <<<"${ROWGROUPS_OUT}"; then
+  echo "FAIL -r output should not include per-column details (encodings/codec)"
+  printf '%s\n' "${ROWGROUPS_OUT}"
+  exit 1
+fi
+echo "OK -r output verified"
+
+# -d / --row-group-details: encodings, codec, stats
+DETAILS_OUT=$({{ICE_CLI}} --config {{CLI_CONFIG}} describe-parquet -d "${TARGET}")
+for expected in "rowGroups:" "encodings:" "codec:" "valueCount:" "path:"; do
+  if ! grep -qF "${expected}" <<<"${DETAILS_OUT}"; then
+    echo "FAIL -d output missing '${expected}'"
+    printf '%s\n' "${DETAILS_OUT}"
+    exit 1
+  fi
+done
+if ! grep -q "nulls:" <<<"${DETAILS_OUT}"; then
+  echo "FAIL -d output missing column stats ('nulls:')"
+  printf '%s\n' "${DETAILS_OUT}"
+  exit 1
+fi
+echo "OK -d output verified"
+
+# -a / --all: every section present
+ALL_OUT=$({{ICE_CLI}} --config {{CLI_CONFIG}} describe-parquet -a "${TARGET}")
+for expected in "summary:" "columns:" "rowGroups:" "encodings:" "codec:" "nulls:"; do
+  if ! grep -qF "${expected}" <<<"${ALL_OUT}"; then
+    echo "FAIL -a output missing '${expected}'"
+    printf '%s\n' "${ALL_OUT}"
+    exit 1
+  fi
+done
+echo "OK -a output verified"
+
+# --json: JSON output instead of YAML
+JSON_OUT=$({{ICE_CLI}} --config {{CLI_CONFIG}} describe-parquet -a --json "${TARGET}")
+if ! grep -q '^{' <<<"${JSON_OUT}"; then
+  echo "FAIL --json output is not a JSON object"
+  printf '%s\n' "${JSON_OUT}"
+  exit 1
+fi
+for expected in '"summary"' '"columns"' '"rowGroups"' "\"rows\":${EXPECTED_ROWS}" "\"columnCount\":${EXPECTED_COLUMN_COUNT}"; do
+  if ! grep -qF "${expected}" <<<"${JSON_OUT}"; then
+    echo "FAIL --json output missing '${expected}'"
+    printf '%s\n' "${JSON_OUT}"
+    exit 1
+  fi
+done
+if grep -qE "^summary:" <<<"${JSON_OUT}"; then
+  echo "FAIL --json output looks like YAML"
+  printf '%s\n' "${JSON_OUT}"
+  exit 1
+fi
+echo "OK --json output verified"
+
+# Combined --summary --columns: both sections, no rowGroups
+COMBINED_OUT=$({{ICE_CLI}} --config {{CLI_CONFIG}} describe-parquet -s --columns "${TARGET}")
+if ! grep -q "^summary:" <<<"${COMBINED_OUT}" || ! grep -q "^columns:" <<<"${COMBINED_OUT}"; then
+  echo "FAIL combined -s --columns missing summary or columns section"
+  printf '%s\n' "${COMBINED_OUT}"
+  exit 1
+fi
+if grep -q "^rowGroups:" <<<"${COMBINED_OUT}"; then
+  echo "FAIL combined -s --columns should not include rowGroups section"
+  printf '%s\n' "${COMBINED_OUT}"
+  exit 1
+fi
+echo "OK Combined --summary --columns output verified"
+
+# Missing file should fail
+if {{ICE_CLI}} --config {{CLI_CONFIG}} describe-parquet "file://${SCENARIO_DIR}/does-not-exist.parquet" >/dev/null 2>&1; then
+  echo "FAIL describe-parquet should fail on missing file"
+  exit 1
+fi
+echo "OK Missing file rejected"
+
+echo "Describe-parquet test completed successfully"

ice-rest-catalog/src/test/resources/scenarios/describe-parquet/scenario.yaml

@@ -0,0 +1,10 @@
+name: "Describe parquet metadata"
+description: "Tests describe-parquet against a local iris parquet file across all option flags and both YAML and JSON output formats"
+
+catalogConfig:
+  warehouse: "s3://test-bucket/warehouse"
+
+env:
+  INPUT_FILE: "input.parquet"
+  EXPECTED_ROWS: "150"
+  EXPECTED_COLUMN_COUNT: "5"