I am a user of a SGE based cluster in my university. I had launched a number of jobs each requiring 8 threads. However, once I launched the jobs, I am unable to ssh into the nodes that are running them as I was able to do before I launched the jobs.
Can someone provide some insight as to why this might be happening?
What sort of jobs are you running? Can you post the job arguments you are using as well?
In general I can't see how running jobs will limit being able to log in via ssh, but we'll see. One thought is that if you are using ssh connections within your job script, maybe you need to increase the max number of ssh sessions in your sshd config file.