ambari2.8.0+bigtop3.2.0发行版大数据平台编译指南

前言

ambari和bigtop联合的第一个发行版终于出来了！！！这是在HDP、CDH大数据平台闭源后的第一个开源免费发行版的大数据平台。下面为大家详细介绍Centos7下的编译方法。

组件版本介绍

编译思路

ambari+bigtop并不是打包在一起的，分别对应三个项目：ambari、ambari-metrics、bigtop。所以要分别编译这三个项目，最后将编译好的包提取到一起做成镜像源。另外，bigtop3.2.0不是所有组件都适配了ambari，只需编译上面表格所适配的组件即可。

编译环境

编译环境的准备工作在之前文章里已经写过，请参考《编译环境准备》进行配置；编译时需要星际网络，梯子需要自己搞定。

编译方法

ambari编译步骤

#git克隆ambari源代码
git clone https://github.com/apache/ambari.git

#进入ambari根目录
cd ambari

#切换到2.8分支
git checkout -b branch-2.8 origin/branch-2.8

#开始编译
mvn clean install rpm:rpm -DskipTests

ambari-metrics编译步骤

#git克隆ambari-metrics
git clone https://github.com/apache/ambari-metrics.git

#进入ambari-metrics根目录
cd ambari-metrics

#切换到3.0分支
git checkout -b branch-3.0 origin/branch-3.0

#提前下载编译时需要的4个tar包--有利于编译加速
wget http://repo.bigtop.apache.org.s3.amazonaws.com/bigtop-stack-binary/3.2.0/centos-7/x86_64/hbase-2.4.13-bin.tar.gz
wget http://repo.bigtop.apache.org.s3.amazonaws.com/bigtop-stack-binary/3.2.0/centos-7/x86_64/hadoop-3.3.4.tar.gz
wget https://dl.grafana.com/oss/release/grafana-9.3.2.linux-amd64.tar.gz
wget http://repo.bigtop.apache.org.s3.amazonaws.com/bigtop-stack-binary/3.2.0/centos-7/x86_64/phoenix-hbase-2.4-5.1.2-bin.tar.gz

#修改ambari-metrics/pom.xml文件，将以上4个tar包的url修改成刚才下载好的本地路径
   file://{下载路径}/hbase-2.4.13-bin.tar.gz
   file://{下载路径}/hadoop-3.3.4.tar.gz
   file://{下载路径}/grafana-9.3.2.linux-amd64.tar.gz
   file://{下载路径}/phoenix-hbase-2.4-5.1.2-bin.tar.gz

#编译
mvn clean install -DskipTests -Dbuild-rpm

bigtop编译步骤

#git克隆bigtop源代码
git clone https://github.com/apache/bigtop.git

#进入bigtop根目录
cd bigtop

#切换到3.2分支
git checkout -b branch-3.2 origin/branch-3.2

#修改bigtop/bigtop.bom配置 有两处要修改
#1.修改镜像源为国内镜像源 103、104行
    APACHE_MIRROR = "https://repo.huaweicloud.com/apache"
    APACHE_ARCHIVE = "https://mirrors.aliyun.com/apache"
#2.解开bigtop-select组件的注释 删除273、281行

#安装组件编译所需的依赖
#1.hadoop依赖
yum -y install fuse-devel cmake cmake3 lzo-devel openssl-devel protobuf* cyrus-* 
cp /usr/bin/cmake3 /usr/bin/cmake
#2.zookeeper依赖
yum -y install cppunit-devel
#3.spark依赖
yum -y install R* harfbuzz-devel fribidi-devel libcurl-devel libxml2-devel freetype-devel libpng-devel libtiff-devel libjpeg-turbo-devel pandoc* libgit2-devel
Rscript -e "install.packages(c('knitr', 'rmarkdown', 'devtools', 'testthat', 'e1071', 'survival'), repos='http://mirrors.tuna.tsinghua.edu.cn/CRAN/')"

#修改部分组件源代码
#1.先下载
./gradlew tez-download zeppelin-download flink-download
#2.进入下载目录
cd dl
#3.解压这3个tar
tar -zxvf flink-1.15.3.tar.gz
tar -zxvf apache-tez-0.10.1-src.tar.gz
tar -zxvf zeppelin-0.10.1.tar.gz
#4.修改flink
vi flink-1.15.0/flink-runtime-web/pom.xml ​
在275行 nodeVersion改为v12.22.1​
在276行 npmVersion改为6.14.12
#5.修改tez
vi apache-tez-0.10.1-src/tez-ui/pom.xml 
在37行 allow-root-build改为--allow-root=true
#6.修改zeppelin
vi zeppelin-0.10.1/pom.xml 
在209行plugin.gitcommitid.useNativeGit改为true
vi zeppelin-0.10.1/spark/pom.xml
在50行spark.src.download.url改为https://repo.huaweicloud.com/apache/spark/${spark.archive}/${spark.archive}.tgz
在53行spark.bin.download.url改为https://repo.huaweicloud.com/apache/spark/${spark.archive}/${spark.archive}-bin-without-hadoop.tgz
vi zeppelin-0.10.1/rlang/pom.xml
在41行spark.src.download.url改为https://repo.huaweicloud.com/apache/spark/${spark.archive}/${spark.archive}.tgz
在44行spark.bin.download.url改为https://repo.huaweicloud.com/apache/spark/${spark.archive}/${spark.archive}-bin-without-hadoop.tgz
vi zeppelin-0.10.1/flink/flink-scala-parent/pom.xml
在45行flink.bin.download.url改为https://repo.huaweicloud.com/apache/flink/flink-${flink.version}/flink-${flink.version}-bin-scala_${flink.scala.binary.version}.tgz
#7.重新打包这3个tar
tar -zcvf flink-1.15.3.tar.gz flink-1.15.3
tar -zcvf apache-tez-0.10.1-src.tar.gz apache-tez-0.10.1-src
tar -zcvf zeppelin-0.10.1.tar.gz zeppelin-0.10.1

#回到bigtop根目录
cd ../

#编译-预计需要一个半小时以上 
./gradlew allclean bigtop-groovy-rpm bigtop-jsvc-rpm bigtop-select-rpm bigtop-utils-rpm flink-rpm hadoop-rpm hbase-rpm hive-rpm kafka-rpm solr-rpm spark-rpm tez-rpm zeppelin-rpm zookeeper-rpm -Dbuildwithdeps=true -PparentDir=/usr/bigtop -PpkgSuffix

制作镜像

#创建bigdatarepo文件夹-路径随意放
mkdir -p bigdatarepo

#将ambari包拷贝
mkdir -p bigdatarepo/ambari
cp ambari/ambari-server/target/rpm/ambari-server/RPMS/x86_64/ambari-server-2.8.0.0-0.x86_64.rpm bigdatarepo/ambari/
cp ambari/ambari-agent/target/rpm/ambari-agent/RPMS/x86_64/ambari-agent-2.8.0.0-0.x86_64.rpm bigdatarepo/ambari/

#将ambari-metrics包拷贝
mkdir -p bigdatarepo/ambari-metrics
cp ambari-metrics/ambari-metrics-assembly/target/rpm/ambari-metrics-collector/RPMS/x86_64/ambari-metrics-collector-3.0.1-1.x86_64.rpm bigdatarepo/ambari-metrics/
cp ambari-metrics/ambari-metrics-assembly/target/rpm/ambari-metrics-grafana/RPMS/x86_64/ambari-metrics-grafana-3.0.1-1.x86_64.rpm bigdatarepo/ambari-metrics/
cp ambari-metrics/ambari-metrics-assembly/target/rpm/ambari-metrics-hadoop-sink/RPMS/x86_64/ambari-metrics-hadoop-sink-3.0.1-1.x86_64.rpm bigdatarepo/ambari-metrics/
cp ambari-metrics/ambari-metrics-assembly/target/rpm/ambari-metrics-monitor/RPMS/x86_64/ambari-metrics-monitor-3.0.1-1.x86_64.rpm bigdatarepo/ambari-metrics/

#将bigtop包拷贝
cp -r bigtop/output/* bigdatarepo/

#制作镜像源
createrepo bigdatarepo/

写在最后

按以上步骤执行，就可以成功编译了，不过一定要注意网络问题。如果觉得编译太麻烦，也可以下载我编译好的包，欢迎一起交流。

echo "编-译-好-的-包-放-在-群-文-件-里-了"
echo "欢-迎-加-Q-Q-群-进-行-交-流"
echo "7-2-2-0-1-4-9-1-2"