/Apache Hadoop Yarn
Data Analysis Tools
Apache Hadoop Yarn Monitoring
YARN governs and orchestrates the processing of data in Hadoop. The YARN cluster architecture is a master-slave cluster framework like HDFS, with a master node daemon called the ResourceManager and one or more slave node daemons called NodeManagers running on worker or slave nodes in the cluster. Moein monitoring system is capable of monitoring YARN Cluster. Performance metrics of ResourceManager and NodeManagers, and their JVM metrics. The following is a list of the performance metrics of Apache YARN:
ResourceManager Metrics:
JVM Metrics:
A complete description of the JVM-related metrics and parameters is provided in the JVM object.
Cluster Metrics:
- ResourceManager Host Name
- Number Of Active NodeManagers
- Number Of Decommissioned NodeManagers
- Number Of Lost NodeManagers
- Number Of Unhealthy NodeManagers
- Number Of Rebooted NodeManagers
- Number Of Launched AMs
- AMs Average Launch Time
- Number Of Registered AMs
- AMs Average Register Time
- Submitted Applications
- Running Applications
- Pending Applications
- Completed Applications
- Killed Applications
Live Nodes Metrics:
- NodeManager Host Name
- NodeManager Rack Name
- NodeManager State
- NodeManager ID
- NodeManager HTTP Address
- NodeManager Last Health Update
- NodeManager Health Report
- NodeManager Version
- NodeManager Number Of Containers
- NodeManager Used Memory
- NodeManager Free Memory Space
- NodeManager Used Memory Percentage
General information from ResourceManager:
- ResourceManager State
- Security Activation Status
- Host Name
- Port
General information from Queues:
- Queue Name
- Queue Host Name
- Running Applications With Elapsed Time Less Than 60 Min
- Running Applications With Elapsed Time Between 60 And 300 Min
- Running Applications With Elapsed Time Between 300 And 1440 Min
- Running Applications With Elapsed Time More Than 1440 Min
- Submitted Applications
- Running Applications
- Pending Applications
- Completed Applications
- Killed Applications
- Allocated Memory
- Allocated CPU
- Allocated Containers
- Total Allocated Containers
- Number Of Rack Local Containers Allocated
- Number Of Off Switch Containers Allocated
- Number Of Released Containers
- Available Memory
- Available CPU
- Pending Requested Memory
- Pending Requested CPU
- Number Of Pending Containers
- Reserved Memory
- Reserved CPU
- Reserved Containers
- Active Users
- Active Applications
NodeManager Metrics:
JVM Metrics:
A complete description of the JVM-related metrics and parameters is provided in the JVM object.
NodeManager Metrics:
- Launched Containers
- Completed Containers
- Failed Containers
- Killed Containers
- Initializing Containers
- Running Containers
- Allocated Memory
- Allocated Containers
- Available Memory
- Used VCores
- Available VCores
Communication Protocols:
- JMX
Data Analysis Tools
Behpaya Co. All rights reserved