Explanation
This is given with respect to Hortonworks Data Platform (HDP). Ambari eliminates the need for the manual tasks that used to watch over Hadoop operations. It gives a simple and secure platform for provisioning, managing, and monitoring HDP deployments. Ambari is an easy to use Hadoop management UI and is solidly backed by REST APIs. The benefits of using Apache Ambari are mentioned below.
Simplified installation, configuration, and management of the Hadoop cluster: Ambari can efficiently create Hadoop clusters at scale. Its wizard-driven approach lets the configuration be automated as per the environment so that the performance is optimal. Master–slave and client components are assigned to configuring services. It is also used to install, start, and test the cluster.
Configuration blueprints give recommendations to those seeking a hands-on approach. The blueprint of an ideal cluster is stored. How it is provisioned is clearly traced. This is then used to automate the creation of successive clusters without any user interaction. Blueprints also preserve and ensure the application of best practices across different environments.
Ambari provides a rolling upgrade feature where running clusters can be updated on the go with maintenance releases and feature-bearing releases, and therefore there is no unnecessary downtime. When there are large clusters involved, rolling updates are simply not possible, in which case express updates are used. Unlike the previous case, here, there is downtime involved but is minimum as when the update is manual. Both rolling and express updates are free of manual updates.
Centralized security and application: The complexity of cluster security configuration and administration is greatly reduced by Ambari which is among the components of the Hadoop ecosystem. The tool also helps the automated setup of the advanced security constructs like Kerberos and Ranger.
Complete visibility to your cluster’s health: Through this tool, you can monitor your cluster’s health and availability. An easily customized web-based dashboard has metrics that give status information for each service in the cluster like HDFS, YARN, and HBase. The tool also helps with garnering and visualizing critical operational metrics for troubleshooting and analysis. Ambari predefines alerts that are integrated with the existing enterprise monitoring tools that monitor cluster components and hosts as per the specified check intervals. Through the browser interface, users can browse alerts for their clusters, search, and filter alerts. They can also view and modify alert properties and alert instances.
Metrics visualization and dashboarding: It provides a scalable low-latency storage system for Hadoop component metrics. Picking the metrics of Hadoop which truly matter requires considerable expertise and understanding of how the components work with each other. Grafana is a leading graph and dashboard builder that simplifies the metrics reviewing process. This is included with Ambari Metrics, along with HDP.
Extensibility and customization: Ambari lets a developer work on Hadoop gracefully in his/her enterprise setup. Ambari leverages the large innovative community which improves upon the tool and it also eliminates vendor lock-in. REST APIs along with Ambari Stacks and Views allows extensive flexibility for customization of HDP implementation.
Ambari Stacks wraps the life cycle control layer used to rationalize operations over a broad set of services. This includes a consistent approach that Ambari uses to manage different types of services like install, start, configure, status, and stop. When provisioning, cluster install experience is rationalized across a set of services by Stacks technology. A natural extension point for operators is provided by Stacks to plug in newly created services that can perform alongside Hadoop.
Third parties can plug in their views through Ambari Views. A view is an application that is deployed into an Ambari container where it offers UI capabilities to be plugged in to give out custom visualization, management, and monitoring features.
Raj Kumar