Slow to execute cmd on the Endeavour cluster

Hi, I’m running jobs on the Endeavour cluster. Whenever I put one cmd (such as top, ls, et…), it takes several seconds to give output, especially after logging in to the compute node. Most of my jobs are in the ‘D’ (Uninterruptible sleep) state. Appreciate your help in fixing this. Thanks in advance.

Is this in a /project directory? The /project file system is slow right now because of metadata issues.

Not only in /project directory, even in the /home1 directory. an obvious lag on the Endeavour compute node.

1 Like