All the Jenkins instances we have that uses JClouds plugin with CloudStack fails to reconnect to generated slave instances whenever Jenkins restart.

I see the following error message for the slave nodes:

This node is offline because Jenkins failed to launch the slave agent on it.

If I try to look at the slave node log, it shows something like the following with a spinner image constantly spinning at the end of the log:

No public addresses found, so using private address.
Connecting to 10.10.51.72 on port 22. 
Connected via SSH.

Looks like the problem is related to http://code.google.com/p/jclouds/issues/detail?id=920

For our Jenkins jobs, we use the jenkins-maven-plugin for generating and maintaining the jobs. This requires us to restart Jenkins quite often. Our build times takes approximately three times longer to build on a clean slave than on a dirty one (due to downloading Maven dependencies). It would be great if we can reuse the existing slave instances.