Can we able to kill particular container in the cluster using yarn command or REST API?

[#1]

@Ram D

To kill an Application you can also use the Application State by using a PUT operation to set the application state to KILLED. For example:

curl -v -X PUT -d '{"state": "KILLED"}''http://localhost:8088/ws/v1/cluster/apps/application_1409421698529_0012'

Hope that helps

[#2]

@Geoffrey Shelton OkotI am asking to kill the container directly. I am aware of killing application. Thank you.

[#3]

@Geoffrey Shelton Okot we are writing a multi-node application that "heals" itself when a node is lost. In order to test the "healing" process we need to kill one of the node containers. Is there a way to do this? Is there any application in the Hadoop ecosystem that does this?

We are running our packages through slider in hadoop cluster as long running jobs.

[#4]

use the slider kill-container command; it's how we test slider apps resilience to failure. There's also a built in chaos-monkey in slider; you can configure the AM to randomly kill containers (and/or its own). See Configuring the Chaos Monkey

[#5 · Accepted solution]

@Ram D

Did you take a look at this.

https://slider.incubator.apache.org/docs/manpage.html

Commands for testing

These operations are here primarily for testing.

kill-container <name> --id container-id

Kill a YARN container belong to the application. This is useful primarily for testing the resilience to failures.

Container IDs can be determined from the application instance status JSON document.

[#6]

@Shivaji Thank you.