Monitoring Hadoop 2 Cluster

A very important aspect of cluster management is monitoring its health, progress status etc. Ankush provides an exhaustive graphical representation and reporting system of varied aspects. There are numerous generic features / aspects that are covered across technologies as follows:

• Tiles: At very top various tiles are shown which can be classified into different categories like alerts, warning & information. Some of the tiles are clickable & opens up corresponding page related to it. In order to view Hadoop 2 Technology specific tiles, refer to Tiles Summary section.

• Trend utilization of CPU, Memory, Network, Load and Packet

• Heat Map: This section provides CPU / Memory utilization HeatMap of cluster nodes. Each block corresponds to a node in the cluster. Color of individual block changes independently as per the value of utilization on the corresponding node. By default CPU utilization is shown. By clicking on required metric “CPU” or “Memory” its utilization heat map can be viewed.

• The node IP to which that HeatMap block corresponds can be identified by resting mouse pointer on it.

• Clicking on HeatMap of a particular node opens up node details page for that node.

• EcoSystem: This section provides details about various ecosystem components installed on cluster. Individual component details can be viewed by clicking on “>” against each component. This normally includes information related to Nodes, Configuration, Events, Logs and Audit Trails. However, the reporting system is further enhanced and provide lots of information about other components as well like Job Monitoring for a Hadoop 2 cluster etc.

• Events: This summarizes various events occurred on the cluster.

• Logs: This functionality allows user to view or download logs of a cluster.

• Audit Trail: This functionality helps to track changes in the configuration.

Please refer to Cluster Monitoring general sections for more information.

![] (https://raw.githubusercontent.com/wiki/impetus-opensource/ankush/images/hadoop/Image20.png) Figure 20: Cluster Monitoring - Generic features

Note: Each technology contains Cluster details page which displays Node list section through the user can further navigate to Node details page. Refer to individual sections for more details.