Hi, I've just found the reason. From what I can see, on the server the process received SIGHUP right after fork. It had no time to switch to new session and detach from controlling terminal. And to work around that one might add sleep to the end: ssh user@xxxxxxxxxxx -t '. ~/.bash_profile && rvm 2.3.1 && cd 1 && ./1.rb && sleep 1' or wait for pid file to be created, whatever. Regards, Yuri _______________________________________________ CentOS mailing list CentOS@xxxxxxxxxx https://lists.centos.org/mailman/listinfo/centos