Support Questions

Find answers, ask questions, and share your expertise

CDH5.4.1: Problem about Yarn NodeManager chart metric

avatar
Explorer

There is a cluster of 15 nodes, several of which Node's chart , the y-axis has a negative case,

How dose this happen?  is there a solution?

allocate container.PNGallocated Memory.PNGrealtime_container_running.PNG

 

1 ACCEPTED SOLUTION

avatar
Explorer

I found the answer,the problem is a bug.

apache jira issue:

https://issues.apache.org/jira/browse/YARN-4408

CDH repaire version:

https://www.cloudera.com/documentation/enterprise/release-notes/topics/cdh_rn_fixed_in_55.html

 

 

Yarn for the container management, through the life-cycle approach, we can know the current operating status of each container, after cloudera manager statistical analysis, but also become a chart metric numerical resources. Therefore, we think this is because there is a problem with container life cycle processing.

View solution in original post

1 REPLY 1

avatar
Explorer

I found the answer,the problem is a bug.

apache jira issue:

https://issues.apache.org/jira/browse/YARN-4408

CDH repaire version:

https://www.cloudera.com/documentation/enterprise/release-notes/topics/cdh_rn_fixed_in_55.html

 

 

Yarn for the container management, through the life-cycle approach, we can know the current operating status of each container, after cloudera manager statistical analysis, but also become a chart metric numerical resources. Therefore, we think this is because there is a problem with container life cycle processing.