Ah seems like it is a known issue with ntfs… ugh:
Seems like I might need to do some fiddling around with /etc/fstab/, but have a starting point at least…
Oli
Update:
The second link above, followed by chown-ing and chmod-ing the cryosparc database directory ultimately fixed the issue. Initially it was still giving the same database error, so I nuked everything in the database directory. Then, cryosparc could start successfully, but I couldn’t log in (it couldn’t find me as a user). So I deleted the master, worker, and database directories, and performed a fresh install. That terminated with the error below. Turned out this was because a cryosparcm process was still running from the previous install; killing that process and rerunning the install script gave a working installation.
Update 2:
On this system, I have five GPUs, of which I want to use four for compute and one to drive the display. I was able to list the gpus using bin/cryosparcw gpulist, and enable the specific GPUs I wanted by running:
bin/cryosparcw connect --update --master ubuntu --worker ubuntu --gpus 0,1,2,3
