Our Community is getting an upgrade! To get everything ready for the relaunch, we’ll be placing the site in read-only mode starting September 21st.
We really appreciate your understanding while we get things set up behind the scenes. Catch up on all the exciting details about the move here.
Need help or have questions? Drop us a line at [email protected]

Archives of Support Questions (Read Only)

This is an archived board for historical reference. Information and links may no longer be available or relevant
Announcements
This board is archived and read-only for historical reference. To ask a new question, please post a new topic on the appropriate active board.

how to calculate the yarn.nodemanager.resource.cpu-vcores in ambari cluster

avatar

as I read from the doc ,yarn.nodemanager.resource.cpu-vcores are by default ~80% of total vCPUs available on the machine

so just to clear that

in case we have ambari cluster with 3 master machine and 3 workers )( machines , and each worker have 8 core

then the calculate will be 3X8=32 ?

or maybe yarn.nodemanager.resource.cpu-vcores should be calculate per worker machine as: 1X8=8 ?


or else?

Michael-Bronson
1 ACCEPTED SOLUTION

avatar
Super Guru

@Michael Bronson,

It will be 80% of your vCPUs on the machine. If you have 3 machines with 16 cores , then it will be 80% of 16 ie 12.8. It will take the floor value of 12.8 and set it as 12.

You can get the cpus by running the command

cat /proc/cpuinfo | grep processor | wc -l

If there are multiple node managers with different vCPU count , it will take the max of all the vCPUs

Thanks,

Aditya

View solution in original post

1 REPLY 1

avatar
Super Guru

@Michael Bronson,

It will be 80% of your vCPUs on the machine. If you have 3 machines with 16 cores , then it will be 80% of 16 ie 12.8. It will take the floor value of 12.8 and set it as 12.

You can get the cpus by running the command

cat /proc/cpuinfo | grep processor | wc -l

If there are multiple node managers with different vCPU count , it will take the max of all the vCPUs

Thanks,

Aditya