diff --git a/Makefile b/Makefile index 9286ce2..399104d 100644 --- a/Makefile +++ b/Makefile @@ -1,3 +1,5 @@ +DOCKER_NETWORK = hadoop +ENV_FILE = hadoop.env current_branch := $(shell git rev-parse --abbrev-ref HEAD) build: docker build -t bde2020/hadoop-base:$(current_branch) ./base @@ -6,3 +8,13 @@ build: docker build -t bde2020/hadoop-resourcemanager:$(current_branch) ./resourcemanager docker build -t bde2020/hadoop-nodemanager:$(current_branch) ./nodemanager docker build -t bde2020/hadoop-historyserver:$(current_branch) ./historyserver + docker build -t bde2020/hadoop-submit:$(current_branch) ./submit + +wordcount: + docker build -t hadoop-wordcount ./submit + docker run --network ${DOCKER_NETWORK} --env-file ${ENV_FILE} bde2020/hadoop-base:$(current_branch) hdfs dfs -mkdir -p /input/ + docker run --network ${DOCKER_NETWORK} --env-file ${ENV_FILE} bde2020/hadoop-base:$(current_branch) hdfs dfs -copyFromLocal /opt/hadoop-2.8.0/README.txt /input/ + docker run --network ${DOCKER_NETWORK} --env-file ${ENV_FILE} hadoop-wordcount + docker run --network ${DOCKER_NETWORK} --env-file ${ENV_FILE} bde2020/hadoop-base:$(current_branch) hdfs dfs -cat /output/* + docker run --network ${DOCKER_NETWORK} --env-file ${ENV_FILE} bde2020/hadoop-base:$(current_branch) hdfs dfs -rm -r /output + docker run --network ${DOCKER_NETWORK} --env-file ${ENV_FILE} bde2020/hadoop-base:$(current_branch) hdfs dfs -rm -r /input diff --git a/submit/Dockerfile b/submit/Dockerfile new file mode 100644 index 0000000..52b0499 --- /dev/null +++ b/submit/Dockerfile @@ -0,0 +1,13 @@ +FROM bde2020/hadoop-base:1.2.0-hadoop2.8-java8 +MAINTAINER Ivan Ermilov + +COPY WordCount.jar /opt/hadoop/applications/WordCount.jar + +ENV JAR_FILEPATH="/opt/hadoop/applications/WordCount.jar" +ENV CLASS_TO_RUN="WordCount" +ENV PARAMS="/input /output" + +ADD run.sh /run.sh +RUN chmod a+x /run.sh + +CMD ["/run.sh"] diff --git a/submit/WordCount.jar b/submit/WordCount.jar new file mode 100644 index 0000000..7b6a178 Binary files /dev/null and b/submit/WordCount.jar differ diff --git a/submit/run.sh b/submit/run.sh new file mode 100644 index 0000000..5ebe79e --- /dev/null +++ b/submit/run.sh @@ -0,0 +1,3 @@ +#!/bin/bash + +$HADOOP_PREFIX/bin/hadoop jar $JAR_FILEPATH $CLASS_TO_RUN $PARAMS