1. The working principle and structure of Ambari
Apache Ambari is an integrated operation and maintenance management tool for big data platforms, which supports the deployment, management and monitoring of Apache Hadoop clusters. Currently, Ambari supports most components of the Hadoop ecosystem, such as HDFS, MapReduce, Hbase, Spark, Storm, Kafka, Druid, Zookeeper, Hive, Pig, Sqoop, Oozie, etc.
Ambari is an application tool with a distributed architecture. It is mainly composed of three parts, namely Ambari Server, Ambari Agent, Ambari Web and Metrics Collector. Send the running status of each software module on each machine to Ambari Server, and Metrics Collector is used to provide a monitoring information query interface for Ambari Server to query, and Ambari Server will summarize all status information and finally present it on Ambari's web page so that we can understand the various status data of the cluster and perform corresponding management and maintenance.
2. Install and deploy Ambari
1. Preparations before deployment
For the installation of Ambari, there are currently two optional release versions, one is Ambari of Apache, and the other is Ambari of Hortonworks. The rpm package version is provided. Therefore, the installation is simpler, and Hortonworks' Amba is used here