Dev

.gitignore

@@ -0,0 +1,3 @@
+**/build
+.gradle
+.vscode

.lfsconfig

@@ -1,2 +1,2 @@
 [lfs]
-    url = https://github.com/AdvancedJavaLabs/lab4-parallel.git/info/lfs
+    url = https://github.com/AdvancedJavaLabs/lab3-kek-flip.git/info/lfs

docker-compose.yaml

@@ -0,0 +1,9 @@
+services:
+  hadoop:
+    build:
+      context: .
+      dockerfile: hadoop/Dockerfile
+    ports:
+      - "9870:9870"
+    volumes:
+      - ./results:/results

hadoop/Dockerfile

@@ -0,0 +1,66 @@
+# Download hadoop
+FROM eclipse-temurin:17-jdk AS base
+
+ENV HADOOP_VERSION=3.4.2
+RUN wget -O /tmp/hadoop.tar.gz "https://dlcdn.apache.org/hadoop/common/hadoop-$HADOOP_VERSION/hadoop-$HADOOP_VERSION-lean.tar.gz"
+
+
+# Install dependencies
+FROM base AS deps
+
+RUN apt-get update && \
+    apt-get install -y --no-install-recommends ssh pdsh sudo
+
+
+# Install hadoop
+FROM deps AS hadoop-installation
+
+ENV HADOOP_HOME=/opt/hadoop
+ENV HADOOP_CONFIG_DIR=$HADOOP_HOME/etc/hadoop
+ENV PATH=$PATH:$HADOOP_HOME/bin:$HADOOP_HOME/sbin
+ENV HADOOP_USER=hadoop
+ENV HADOOP_UID=1001
+ENV HADOOP_GID=1001
+
+RUN groupadd -g $HADOOP_GID $HADOOP_USER && \
+    useradd -u $HADOOP_UID -g $HADOOP_GID -r -m -s /bin/bash $HADOOP_USER && \
+    usermod -aG sudo $HADOOP_USER && \
+    echo "$HADOOP_USER ALL=(ALL) NOPASSWD:ALL" >> /etc/sudoers
+
+RUN tar -xzf /tmp/hadoop.tar.gz -C /opt && \
+    mv /opt/hadoop-$HADOOP_VERSION /opt/hadoop && \
+    rm /tmp/hadoop.tar.gz
+
+
+# Configure hadoop
+FROM hadoop-installation AS hadoop-configuration
+
+COPY hadoop/core-site.xml $HADOOP_CONFIG_DIR/core-site.xml
+COPY hadoop/hdfs-site.xml $HADOOP_CONFIG_DIR/hdfs-site.xml
+
+
+# Prepare runtime
+FROM hadoop-configuration AS runtime
+
+RUN echo "export JAVA_HOME=$JAVA_HOME" >> $HADOOP_HOME/etc/hadoop/hadoop-env.sh
+RUN echo "export HDFS_NAMENODE_USER=$HADOOP_USER" >> $HADOOP_HOME/etc/hadoop/hadoop-env.sh
+RUN echo "export HDFS_DATANODE_USER=$HADOOP_USER" >> $HADOOP_HOME/etc/hadoop/hadoop-env.sh
+RUN echo "export HDFS_HADOOP_USER=$HADOOP_USER" >> $HADOOP_HOME/etc/hadoop/hadoop-env.sh
+
+ENV ARTIFACT_NAME=sales-analyzer
+ENV JAR_NAME=$ARTIFACT_NAME-1.0.jar
+ENV BUILD_PATH=$ARTIFACT_NAME/build/libs
+COPY $BUILD_PATH/$JAR_NAME .
+
+COPY *.csv .
+
+COPY hadoop/start.sh start.sh
+RUN chmod +x start.sh
+
+USER $HADOOP_USER
+
+RUN ssh-keygen -t rsa -P "" -f ~/.ssh/id_rsa && \
+    cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys && \
+    chmod 600 ~/.ssh/authorized_keys
+
+ENTRYPOINT ["./start.sh"]

hadoop/core-site.xml

@@ -0,0 +1,6 @@
+<configuration>
+    <property>
+        <name>fs.defaultFS</name>
+        <value>hdfs://localhost:9000</value>
+    </property>
+</configuration>

hadoop/hdfs-site.xml

@@ -0,0 +1,10 @@
+<configuration>
+    <property>
+        <name>dfs.replication</name>
+        <value>1</value>
+    </property>
+    <property>
+        <name>dfs.permissions.enabled</name>
+        <value>false</value>
+    </property>
+</configuration>

hadoop/start.sh

@@ -0,0 +1,45 @@
+#!/bin/bash
+set -e
+
+sudo service ssh start
+sudo chmod o+w /results
+
+NAMENODE_DIR=/opt/hadoop/data/namenode
+if [ ! -d "$NAMENODE_DIR/current" ]; then
+    hdfs namenode -format
+fi
+
+start-dfs.sh
+start-yarn.sh
+
+hadoop fs -mkdir /csv
+hadoop fs -put *.csv /csv
+
+MAPRED_NUM_REDUCE_TASKS_SET=(1 3 5 10 15 20)
+SPLIT_MINSIZE_SET=(1024 2048 8192 32768 524288)
+
+for MAPRED_NUM_REDUCE_TASKS in "${MAPRED_NUM_REDUCE_TASKS_SET[@]}"; do
+  for SPLIT_MINSIZE in "${SPLIT_MINSIZE_SET[@]}"; do
+    dir=tasks-$MAPRED_NUM_REDUCE_TASKS\_split-$SPLIT_MINSIZE
+
+    start_time=$(date +%s%N)
+    hadoop jar sales-analyzer-1.0.jar org.itmo.MainJob $MAPRED_NUM_REDUCE_TASKS $SPLIT_MINSIZE out /csv/*.csv
+    end_time=$(date +%s%N)
+
+    duration_ms=$(($((end_time - start_time)) / 1000000))
+
+    if [ ! -d "results/$dir" ]; then
+      mkdir results/$dir
+    fi
+
+    echo $duration_ms > results/$dir/exec_time_ms
+
+    hadoop fs -get out/part-r-00000 results/$dir/data.csv
+    sed -i '1i category,revenue,quantity' results/$dir/data.csv
+
+    hadoop fs -rm -r out
+    hadoop fs -rm -r mapred_out
+  done
+done
+
+#tail -f /dev/null

sales-analyzer/build.gradle.kts

@@ -0,0 +1,21 @@
+plugins {
+    id("java")
+}
+
+java {
+    toolchain {
+        languageVersion.set(JavaLanguageVersion.of(17))
+    }
+}
+
+group = "org.itmo"
+version = "1.0"
+
+repositories {
+    mavenCentral()
+}
+
+dependencies {
+    compileOnly("org.apache.hadoop:hadoop-common:3.4.2")
+    compileOnly("org.apache.hadoop:hadoop-mapreduce-client-core:3.4.2")
+}

sales-analyzer/gradle/wrapper/gradle-wrapper.properties

@@ -0,0 +1,6 @@
+#Sat Dec 13 15:56:51 MSK 2025
+distributionBase=GRADLE_USER_HOME
+distributionPath=wrapper/dists
+distributionUrl=https\://services.gradle.org/distributions/gradle-8.8-bin.zip
+zipStoreBase=GRADLE_USER_HOME
+zipStorePath=wrapper/dists