I switched the job configuration to run into master server... the waiting job is now runing in the master lost (not sure if it will be able to successfully finish, we have had problems running jobs in slots different to hudson-slave1).... I tied to cancel before it started, but it didn't ... trying to cancel it ...
I guess there should be some kind of failure with hudson
servers... hopefully the cause of failing builds... I'll try todo
something as soon as hudson is available...