Hey there. Maybe someone can help me with this. I'm running Debian 9.0 (stretch). I have a few scripts which periodically create and then move a file from my 1TB M.2 drive to a MD RAID1 device comprised of 2x 6TB Seagate IronWolf drives. Full details are here [1].
My disks look something like this: jj5@tact:~$ df -h Filesystem Size Used Avail Use% Mounted on /dev/md0 469G 86G 378G 19% / /dev/md1 5.5T 4.0T 1.6T 73% /data /dev/nvme0n1p1 931G 78G 853G 9% /data/fast An example of the move command which locks everything up is: mv /data/fast/temp/blob-tact.tgz \ /data/temp/blob-tact.tgz.8157-12003.tmp So while this mv command is being processed my mouse doesn't respond, my KDE Plasma session locks up, everything grinds to a halt. There seems to be plenty of RAM available, and plenty of CPU. So must be some I/O thing..? Anyway, I'm a bit stumped. I haven't tried running the mv command with nice, because it doesn't seem to be a CPU thing... but maybe I should..? Anyone got any ideas about how to keep my system responsive while the mv command is processed? Regards, John Elliot V [1] https://www.jj5.net/sixsigma/Workstation#Disk -- E: j...@jj5.net P: +61 4 3505 7839 W: https://www.jj5.net/
<<attachment: jj5.vcf>>