After I figured out how to turn on more logging messages, I got this:

could not check C:\Users\jenkins\slave\workspace\r-WGI6TZZNQOBAAHSSLVRBWLNOVFN3FM5CSX5AJTOHIS4QEFKE2BPA@2
java.io.IOException: Remote call on Channel to /172.17.xxx.xxx failed
	at hudson.remoting.Channel.call(Channel.java:830)
	at org.jenkinsci.plugins.durabletask.ProcessLiveness._isAlive(ProcessLiveness.java:77)
	at org.jenkinsci.plugins.durabletask.ProcessLiveness.isAlive(ProcessLiveness.java:59)
	at org.jenkinsci.plugins.durabletask.BourneShellScript$ShellController.exitStatus(BourneShellScript.java:198)
	at org.jenkinsci.plugins.workflow.steps.durable_task.DurableTaskStep$Execution.check(DurableTaskStep.java:307)
	at org.jenkinsci.plugins.workflow.steps.durable_task.DurableTaskStep$Execution.run(DurableTaskStep.java:276)
	at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
	at java.util.concurrent.FutureTask.run(FutureTask.java:266)
	at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(ScheduledThreadPoolExecutor.java:180)
	at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:293)
	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
	at java.lang.Thread.run(Thread.java:745)
Caused by: java.lang.NoClassDefFoundError: Could not initialize class org.jenkinsci.plugins.durabletask.ProcessLiveness$LibC
	at org.jenkinsci.plugins.durabletask.ProcessLiveness$Liveness.call(ProcessLiveness.java:98)
	at org.jenkinsci.plugins.durabletask.ProcessLiveness$Liveness.call(ProcessLiveness.java:91)
	at hudson.remoting.UserRequest.perform(UserRequest.java:153)
	at hudson.remoting.UserRequest.perform(UserRequest.java:50)
	at hudson.remoting.Request$2.run(Request.java:336)
	at hudson.remoting.InterceptingExecutorService$1.call(InterceptingExecutorService.java:68)
	at java.util.concurrent.FutureTask.run(FutureTask.java:266)
	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
	at hudson.remoting.Engine$1$1.run(Engine.java:94)
	at java.lang.Thread.run(Thread.java:745)
	at ......remote call to Channel to /172.17.xxx.xxx(Native Method)
	at hudson.remoting.Channel.attachCallSiteStackTrace(Channel.java:1537)
	at hudson.remoting.UserResponse.retrieve(UserRequest.java:253)
	at hudson.remoting.Channel.call(Channel.java:822)
	... 12 more

I've not figured out how to turn on the additional logging? How did you accomplish this? I am curious it the issue I have is the same, or not.

I run running a linux master and agents. My error is very similar.

Cannot contact DevelopmentBuild1_whyme: java.io.IOException: Remote call on DevelopmentBuild1_whyme failed.

My problem started occurring on 2 different Jenkins masters with different agents, and on different machines. The problem seemed to have coincided with plugin updates.

https://wiki.jenkins-ci.org/display/JENKINS/Logging

Thanks for the link!

Thanks to the logging, I have confirmed that my error is the same error that is documented in this issue.. 

We are seeing the same issue with MultibranchPipeline jobs with Windows Server 2012 slaves. Rolling back a number of Pipeline/Workflow plugins solved the issue for us. We were not able to identify exactly which of the pipeline related plugins introduced this issue, but since they are very interdependent it would not help us much anyway.

We are also stuck with an old version of Blue Ocean since it requires a newer version of the Pipeline plugin which would introduce this issue again.

Below is the combination of Pipeline related plugins which has not yet introduce these error messages for us (... java.io.IOException: Remote call on Channel ...):

pipeline-build-step:2.4
pipeline-github-lib:1.0
pipeline-graph-analysis:1.3
pipeline-input-step:2.5
pipeline-milestone-step:1.3
pipeline-model-api:1.0.1
pipeline-model-declarative-agent:1.0.1
pipeline-model-definition:1.0.1
pipeline-multibranch-defaults:1.1
pipeline-rest-api:2.4
pipeline-stage-step:2.2
plain-credentials:1.4
pipeline-stage-tags-metadata:1.0.1
pipeline-stage-view:2.4
workflow-aggregator:2.5
workflow-api:2.10
workflow-basic-steps:2.4
workflow-cps:2.26
workflow-cps-global-lib:2.6
workflow-durable-task-step:2.8
workflow-job:2.10
workflow-multibranch:2.12
workflow-remote-loader:1.4
workflow-scm-step:2.3
workflow-step-api:2.9
workflow-support:2.12

Downgrade of plugin "Pipeline: Nodes and Processes" from v.2.9 to v.2.8 has fixed this problem (at least in my environment)

Thanks Ivan.... I downgraded that plugin form 2.10 to 2.8, and it has fixed the problem in my case as well. 

I can confirm that a dependency to the "Pipeline: Nodes and Processes" plugin introduced the error, and a downgrade from version 2.10 to 2.8 removed the error message;
Cannot contact <slave>: java.io.IOException: Remote call on Channel to /<ip> failed

One plugin which is dependent on version 2.10 of this plugin is BlueOcean 1.0.0-b25

Not really a regression; the problem was always there, but until recently not reported unless you happened to have enabled a FINE logger in DurableTaskStep.

The problem here is that you trying to run the sh step on Windows. It is only tested on Linux. Use bat. Probably sh could be supported under some configurations, but this would require some further development to disable features which will not run in a non-POSIX system, including the process liveness check, as well as an automated test environment capable of running basic scenarios and detecting regressions.

Well, everything works perfectly with Freestyle jobs under Windows. This essentially means, executing shell (bash) scripts under Windows isn't supported with pipelines? And I guess I cannot execute shell scripts with the bat command? Using batch files isn't an option.

This error is present on linux nodes as well...

The main reason for the update of the "Pipeline Nodes and Processes" plugin to version 2.9 was to include a dependency to the "Durable task" plugin to version 1.13, to solve an issue with Environment variables.

It is a bit ironic then that the second main issue solved in the "Durable task" plugin v1.13 was;
"JENKINS-40225 Replace backslashes when on Cygwin to allow sh to be used.", given the information that Pipeline is not supposed to support the shell on Windows

Installing Blue Ocean, which of course needed to update "Pipeline: Nodes and Processes", and of course the problem has returned..   This is on Linux, not Windows.   Still no resolution to this problem, other than to downgrade, which is really not an option for me any longer. 

There is an ongoing fix for JENKINS-38682, which should resolve this issue as well. https://github.com/jenkinsci/durable-task-plugin/pull/43

Well, everything works perfectly with Freestyle jobs under Windows.

No one actually ferd back - but the reason is that this is 2 different beasts. the sh step in a pipeline will survive a disconnect of the agent (ie start a long running sh step and then restart your master, or pull the network cable on your agent and then reconnect it) . Pipelines support durability so the sh step will continue to run and when the master-agent comms are restored the pipeline will pickup the scripts process, in a freestyle build doing this will cause your build will fail.

This essentially means, executing shell (bash) scripts under Windows isn't supported with pipelines? And I guess I cannot execute shell scripts with the bat command? Using batch files isn't an option.

you can call shell scripts from a bat. if you write the shell script to a file then use bat "sh thescript.sh".

Should have been solved—or, at least, changed!—by JENKINS-47791.