When running a build on an remote executor and the build generates a large checkstyle error log(approx 100K issues) the build task locks up after transferring the files to the master.

The last line in the build log reads:
[CheckStyle] -> 4921 copied, 0 not in workspace, 0 not-found, 0 with I/O error
There is no other indication that any activity is ongoing.

If a time limit is imposed on the build it will get killed after the limit is exceeded.

We did find a work around of placing a checkstyle in its own recordIssues surrounded by 5 second delays, as shown below:

sleep 5
recordIssues enabledForFailure: true, forensicsDisabled: true,  blameDisabled: true, ignoreFailedBuilds: true, tools: [
 checkStyle(pattern: "$ncmDirectory/**/build/reports/checkstyle/**/*.xml")]
 sleep 5

This may be similar to JENKINS-32191