Have you checked if there are open file handles on that workspace?

duplicate to https://issues.jenkins-ci.org/browse/JENKINS-10905 ?

I think this is a generic problem, not exactly same as JENKINS-10905. I have the same problem, my jenkins install will work for a few builds then it will continually fail on svn checkout with this exact same failure. It's really bad because once this failure occurs all of my slaves will have the same failure. Yes, I did make sure that the slaves and master are the same version. My Env is jenkins master (ver 1.409.2) on CentOS 5.6 running in a tomcat 6 container. I have 4 slaves: CetnOS (SSH), Win XP (Win Service), Win 7 (JNLP), and Ubuntu (SSH). I get this failure on all of my slaves and the only way to get the builds working again is by restarting the master. Restarting slaves will not fix the problem.

here's a few posting from the google group:
http://groups.google.com/group/jenkinsci-users/browse_thread/thread/8a783e9865bcff11/e20fe877b71108ce#e20fe877b71108ce
http://groups.google.com/group/jenkinsci-users/browse_thread/thread/b404c8ba4edf9f91/8df98229ff76a6dc?lnk=gst&q=khaido#8df98229ff76a6dc

I think this issue is more than related to JENKINS-10771. It's duplicate.

Bruce, could you try and paste a complete stack trace to check it's the same or a different issue.

We had similar problem with our build slaves when use :Launche slave agents on unix machines via ssh

We solved it by use of : Launch slave via execution of command on Master

Launche command:
ssh -o "LogLevel=ERROR" -o "BatchMode=yes" -o "StrictHostKeyChecking=no" -o "UserKnownHostsFile=/dev/null" -i /home/jenkinsuser/.ssh/jenkinsuser.rsa jenkinsuser@buildslave-hostname "java -jar $jenkinshome/.hudson/war/WEB-INF/slave.jar"

Now slaves are running stable again.

No response from the reporter, closing as a duplicate of JENKINS-10771 as batmat suggested 5 years ago.