I’m running into a persistent issue (there are similar posts to this one) where executing unix commands on the command line lags significantly in the /home1 directory (on the Endeavour cluster if this helps). Were there any metadata issues in the past few days? How can we prevent or bypass these problems in the future?
I sympathize, @jh_445; it can be frustrating when everything slows down without explanation. Then when things go back to normal, you don’t know how long that will be the case. Was the problem really fixed, or will everything grind to a halt again shortly?
A standard practice at larger computing centers (e.g. TACC, ORNL, and ACCESS systems) is to proactively alert users when filesystem issues occur, and also when they are remedied. I suggest that the admins at CARC begin to follow this practice, as filesystem slowdowns happen with some regularity but I only remember seeing one such notice (and it occurred days into a severe issue, and after I submitted tickets documenting it).