[FEDE-5150] Upgrade arrow version from 4.0.0 to 7.0.0

README.md

@@ -17,6 +17,73 @@
   under the License.
 -->
 
+# Siren fork of Arrow
+
+- The properties `drill.enable_unsafe_memory_access` and
+  `arrow.enable_unsafe_memory_access` are prefixed with `siren` and their
+  default value is set to `true`. The first property is deprecated.
+
+- In order to avoid conflict with a version of `netty` used in Elasticsearch, we
+  relocate the netty custom package and dependency in `memory` into a package
+  named `siren`. The relocation is achieved thanks to the maven shade plugin.
+
+- The Siren's fork of `netty` is used in `vector`. This means that `netty`
+  imports in that module need to be prefixed with `siren`.
+
+## Check that Siren version of Netty is used
+- In order to check that Siren version of Netty is being used, 
+  run the unit test `CheckAccessibleTest` in 
+  `https://github.com/sirensolutions/siren-platform/blob/master/core/src/test/java/io/siren/federate/core/common/CheckAccessibleTest.java`.
+- Note: the unit test `CheckAccessibleTest` is currently ignored, please set it again to ignore after running the test.
+  The unit test is ignored because the settings in `CheckAccessibleTest` is not taken into account when the whole unit test suite is run, therefore it fails. 
+  This could be because when the class is loaded, the default settings is used (which is a static block) and the new settings in the `CheckAccessibleTest` is
+  then not applied when the test suit is run.
+
+## Build
+
+To build the `memory`, `format` and `vector` modules:
+
+```sh
+$ cd java
+$ mvn clean package
+```
+
+Because of the default value change of `unsafe_memory_access` property, some
+tests in `vector` fail.
+
+```sh
+mvn -pl memory,memory/memory-core,memory/memory-netty,memory/memory-unsafe,format,vector install -Dsiren.arrow.enable_unsafe_memory_access=false -Dsiren.drill.enable_unsafe_memory_access=false
+```
+
+## Make a new release of Siren's Apache Arrow
+
+- Tests should pass.
+
+- Make a new version:
+
+```sh
+mvn versions:set -DnewVersion=siren-0.14.1-2
+```
+
+- tag the commit for the release
+
+```sh
+git tag --sign siren-0.14.1-2
+````
+
+- Deploy to Siren's artifactory
+
+```sh
+$ mvn deploy -DskipTests=true -P artifactory -Dartifactory_username=<USERNAME> -Dartifactory_password=<PASSWORD>
+```
+
+## Update to a new version of Siren's Apache Arrow
+
+- add `git@github.com:apache/arrow.git` as the `upstream` remote.
+- execute `git fetch --all --tags`
+- create a temporary branch from `siren-changes`
+- rebase against the new tag.
+
 # Apache Arrow
 
 [![Fuzzing Status](https://oss-fuzz-build-logs.storage.googleapis.com/badges/arrow.svg)](https://bugs.chromium.org/p/oss-fuzz/issues/list?sort=-opened&can=1&q=proj:arrow)

java/adapter/orc/pom.xml

@@ -87,7 +87,7 @@
     <parent>
         <groupId>org.apache.arrow</groupId>
         <artifactId>arrow-java-root</artifactId>
-        <version>7.0.0</version>
+        <version>siren-7.0.0-1-SNAPSHOT</version>
         <relativePath>../../pom.xml</relativePath>
     </parent>
 

java/format/pom.xml

@@ -15,7 +15,7 @@
 <parent>
   <artifactId>arrow-java-root</artifactId>
   <groupId>org.apache.arrow</groupId>
-  <version>7.0.0</version>
+  <version>siren-7.0.0-1-SNAPSHOT</version>
 </parent>
 
 <artifactId>arrow-format</artifactId>

java/gandiva/pom.xml

@@ -14,7 +14,7 @@
     <parent>
       <groupId>org.apache.arrow</groupId>
       <artifactId>arrow-java-root</artifactId>
-      <version>7.0.0</version>
+      <version>siren-7.0.0-1-SNAPSHOT</version>
     </parent>
 
     <groupId>org.apache.arrow.gandiva</groupId>

java/memory/memory-core/pom.xml

@@ -13,7 +13,7 @@
   <parent>
     <artifactId>arrow-memory</artifactId>
     <groupId>org.apache.arrow</groupId>
-    <version>7.0.0</version>
+    <version>siren-7.0.0-1-SNAPSHOT</version>
   </parent>
   <modelVersion>4.0.0</modelVersion>
 

java/memory/memory-core/src/main/java/org/apache/arrow/memory/BoundsChecking.java

@@ -34,14 +34,15 @@ public class BoundsChecking {
   static final org.slf4j.Logger logger = org.slf4j.LoggerFactory.getLogger(BoundsChecking.class);
 
   static {
-    String envProperty = System.getenv("ARROW_ENABLE_UNSAFE_MEMORY_ACCESS");
-    String oldProperty = System.getProperty("drill.enable_unsafe_memory_access");
+    String envProperty = System.getenv().getOrDefault("SIREN_ARROW_ENABLE_UNSAFE_MEMORY_ACCESS", "true");
+    String oldProperty = System.getProperty("siren.drill.enable_unsafe_memory_access", "true");
     if (oldProperty != null) {
-      logger.warn("\"drill.enable_unsafe_memory_access\" has been renamed to \"arrow.enable_unsafe_memory_access\"");
-      logger.warn("\"arrow.enable_unsafe_memory_access\" can be set to: " +
-              " true (to not check) or false (to check, default)");
+      logger.warn("\"siren.drill.enable_unsafe_memory_access\" has been renamed to " +
+              "\"siren.arrow.enable_unsafe_memory_access\"");
+      logger.warn("\"siren.arrow.enable_unsafe_memory_access\" can be set to: " +
+              " true (to not check, default) or false (to check)");
     }
-    String newProperty = System.getProperty("arrow.enable_unsafe_memory_access");
+    String newProperty = System.getProperty("siren.arrow.enable_unsafe_memory_access", "true");
 
     // The priority of determining the unsafe flag:
     // 1. The system properties take precedence over the environmental variable.

java/memory/memory-core/src/main/java/org/apache/arrow/memory/util/MemoryUtil.java

@@ -81,7 +81,6 @@ public Object run() {
 
       // get the offset of the address field in a java.nio.Buffer object
       Field addressField = java.nio.Buffer.class.getDeclaredField("address");
-      addressField.setAccessible(true);
       BYTE_BUFFER_ADDRESS_OFFSET = UNSAFE.objectFieldOffset(addressField);
 
       Constructor<?> directBufferConstructor;
@@ -99,10 +98,7 @@ public Object run() {
                   constructor.setAccessible(true);
                   logger.debug("Constructor for direct buffer found and made accessible");
                   return constructor;
-                } catch (NoSuchMethodException e) {
-                  logger.debug("Cannot get constructor for direct buffer allocation", e);
-                  return e;
-                } catch (SecurityException e) {
+                } catch (Exception e) {
                   logger.debug("Cannot get constructor for direct buffer allocation", e);
                   return e;
                 }

java/memory/memory-core/src/test/java/org/apache/arrow/memory/TestBoundaryChecking.java

@@ -55,15 +55,15 @@ private boolean getFlagValue(ClassLoader classLoader) throws Exception {
   }
 
   /**
-   * Ensure the flag for bounds checking is enabled by default.
+   * Siren: Ensure the flag for bounds checking is disabled by default.
    * This will protect users from JVM crashes.
    */
   @Test
   public void testDefaultValue() throws Exception {
     ClassLoader classLoader = copyClassLoader();
     if (classLoader != null) {
       boolean boundsCheckingEnabled = getFlagValue(classLoader);
-      Assert.assertTrue(boundsCheckingEnabled);
+      Assert.assertFalse(boundsCheckingEnabled);
     }
   }
 

java/memory/memory-netty/pom.xml

@@ -13,7 +13,7 @@
   <parent>
     <artifactId>arrow-memory</artifactId>
     <groupId>org.apache.arrow</groupId>
-    <version>7.0.0</version>
+    <version>siren-7.0.0-1-SNAPSHOT</version>
   </parent>
   <modelVersion>4.0.0</modelVersion>
 
@@ -35,16 +35,65 @@
       <groupId>io.netty</groupId>
       <artifactId>netty-common</artifactId>
     </dependency>
+    <!--provided by elasticsearch-->
     <dependency>
       <groupId>org.slf4j</groupId>
       <artifactId>slf4j-api</artifactId>
+      <scope>provided</scope>
     </dependency>
     <dependency>
       <groupId>org.immutables</groupId>
       <artifactId>value</artifactId>
     </dependency>
   </dependencies>
 
+  <build>
+    <plugins>     
+      <plugin>
+        <groupId>org.apache.maven.plugins</groupId>
+        <artifactId>maven-shade-plugin</artifactId>
+        <version>3.1.0</version>
+        <executions>
+          <execution>
+            <phase>package</phase>
+            <goals>
+              <goal>shade</goal>
+            </goals>
+            <configuration>
+              <shadedArtifactAttached>false</shadedArtifactAttached>
+              <dependencyReducedPomLocation>${project.build.directory}/dependency-reduced-pom.xml</dependencyReducedPomLocation>
+              <relocations>
+                <relocation>
+                  <pattern>io.netty</pattern>
+                  <shadedPattern>siren.io.netty</shadedPattern>
+                </relocation>
+              </relocations>
+              <artifactSet>
+                <excludes>
+                  <exclude>org.slf4j</exclude>
+                  <exclude>com.google.code.findbugs</exclude>
+                  <exclude>com.google.guava</exclude>
+                  <exclude>org.apache.arrow:arrow-memory-core:*:*</exclude>
+                </excludes>
+              </artifactSet>
+              <!--to prevent the Invalid signature file error -->
+              <filters>
+                <filter>
+                  <artifact>*:*</artifact>
+                  <excludes>
+                    <exclude>META-INF/*.SF</exclude>
+                    <exclude>META-INF/*.DSA</exclude>
+                    <exclude>META-INF/*.RSA</exclude>
+                  </excludes>
+                </filter>
+              </filters>
+            </configuration>
+          </execution>
+        </executions>
+      </plugin>
+    </plugins>
+  </build>
+
   <profiles>
     <profile>
       <!-- This profile turns on integration testing. It activates the failsafe plugin and will run any tests

java/memory/memory-unsafe/pom.xml

@@ -13,7 +13,7 @@
   <parent>
     <artifactId>arrow-memory</artifactId>
     <groupId>org.apache.arrow</groupId>
-    <version>7.0.0</version>
+    <version>siren-7.0.0-1-SNAPSHOT</version>
   </parent>
   <modelVersion>4.0.0</modelVersion>
 

java/memory/pom.xml

@@ -14,7 +14,7 @@
   <parent>
     <groupId>org.apache.arrow</groupId>
     <artifactId>arrow-java-root</artifactId>
-    <version>7.0.0</version>
+    <version>siren-7.0.0-1-SNAPSHOT</version>
   </parent>
   <artifactId>arrow-memory</artifactId>
   <name>Arrow Memory</name>

java/pom.xml

@@ -20,7 +20,7 @@
 
   <groupId>org.apache.arrow</groupId>
   <artifactId>arrow-java-root</artifactId>
-  <version>7.0.0</version>
+  <version>siren-7.0.0-1-SNAPSHOT</version>
   <packaging>pom</packaging>
 
   <name>Apache Arrow Java Root POM</name>
@@ -33,7 +33,7 @@
     <dep.junit.jupiter.version>5.4.0</dep.junit.jupiter.version>
     <dep.slf4j.version>1.7.25</dep.slf4j.version>
     <dep.guava.version>30.1.1-jre</dep.guava.version>
-    <dep.netty.version>4.1.68.Final</dep.netty.version>
+    <dep.netty.version>siren-4.1.68-1</dep.netty.version>
     <dep.jackson.version>2.11.4</dep.jackson.version>
     <dep.hadoop.version>2.7.1</dep.hadoop.version>
     <dep.fbs.version>1.12.0</dep.fbs.version>
@@ -686,14 +686,14 @@
     <module>format</module>
     <module>memory</module>
     <module>vector</module>
-    <module>tools</module>
-    <module>adapter/jdbc</module>
-    <module>plasma</module>
-    <module>flight</module>
-    <module>performance</module>
-    <module>algorithm</module>
-    <module>adapter/avro</module>
-    <module>compression</module>
+    <!--<module>tools</module>-->
+    <!--<module>adapter/jdbc</module>-->
+    <!--<module>plasma</module>-->
+    <!--<module>flight</module>-->
+    <!--<module>performance</module>-->
+    <!--<module>algorithm</module>-->
+    <!--<module>adapter/avro</module>-->
+    <!--<module>compression</module>-->
   </modules>
 
   <profiles>
@@ -846,6 +846,24 @@
       </reporting>
     </profile>
 
+    <!-- To deploy to Siren artifactory -->
+    <profile>
+      <id>artifactory</id>
+
+      <distributionManagement>
+        <repository>
+          <id>artifactory-releases</id>
+          <name>artifactory-releases</name>
+          <url>${artifactory.url}/libs-release-local</url>
+        </repository>
+        <snapshotRepository>
+          <id>artifactory-snapshots</id>
+          <name>artifactory-snapshots</name>
+          <url>${artifactory.url}/libs-snapshot-local</url>
+        </snapshotRepository>
+      </distributionManagement>
+    </profile>
+
   </profiles>
 
 </project>

java/vector/pom.xml

@@ -14,7 +14,7 @@
   <parent>
     <groupId>org.apache.arrow</groupId>
     <artifactId>arrow-java-root</artifactId>
-    <version>7.0.0</version>
+    <version>siren-7.0.0-1-SNAPSHOT</version>
   </parent>
   <artifactId>arrow-vector</artifactId>
   <name>Arrow Vectors</name>
@@ -53,7 +53,6 @@
       <groupId>org.apache.arrow</groupId>
       <artifactId>arrow-memory-netty</artifactId>
       <version>${project.version}</version>
-      <scope>test</scope>
     </dependency>
     <dependency>
       <groupId>org.apache.arrow</groupId>
@@ -121,12 +120,6 @@
             <goals>
               <goal>test</goal>
             </goals>
-            <configuration>
-              <classpathDependencyExcludes>
-                <classpathDependencyExclude>org.apache.arrow:arrow-memory-netty</classpathDependencyExclude>
-              </classpathDependencyExcludes>
-              <reportNameSuffix>netty</reportNameSuffix>
-            </configuration>
           </execution>
         </executions>
       </plugin>

java/vector/src/main/codegen/includes/vv_imports.ftl

@@ -20,6 +20,8 @@ import static org.apache.arrow.util.Preconditions.checkState;
 
 import com.google.flatbuffers.FlatBufferBuilder;
 
+import siren.io.netty.buffer.*;
+
 import org.apache.arrow.memory.*;
 import org.apache.arrow.util.Preconditions;
 import org.apache.arrow.vector.types.Types;

java/vector/src/main/java/org/apache/arrow/vector/BaseFixedWidthVector.java

@@ -37,7 +37,7 @@
 import org.apache.arrow.vector.util.OversizedAllocationException;
 import org.apache.arrow.vector.util.TransferPair;
 
-import io.netty.util.internal.PlatformDependent;
+import siren.io.netty.util.internal.PlatformDependent;
 
 /**
  * BaseFixedWidthVector provides an abstract interface for

java/vector/src/main/java/org/apache/arrow/vector/BitVector.java

@@ -119,6 +119,11 @@ public void setInitialCapacity(int valueCount) {
     lastValueCapacity = valueCount;
   }
 
+  /**
+   * Get the current value capacity for the vector.
+   *
+   * @return number of elements that vector can hold.
+   */
   @Override
   protected int getValueBufferValueCapacity() {
     return capAtMaxInt(valueBuffer.capacity() * 8);