Many questions have been asked & answered about issues with mount.ntfs-3g's CPU usage; typically, the solution offered is to add big_writes to the mount options (e.g. here), or to simply say "it's an inefficient driver, so poor performance is par for the course."
In my case, I've been happily running nearly everything off a shared ntfs partition on a dual-boot machine for some time. Virtually all my data is read/written directly from that partition, and I've never had any issues with CPU.
However, I recently started working with a large & fairly complex web project (node/typescript/react, & using mongo+postgres+redis+localstack). Whenever the server is run from the ntfs partition, mount.ntfs-3g continuously uses >20% CPU. This happens even when the site is not being accessed. Running it from btrfs uses no CPU whatsoever.
So my question: how can I debug what part of the environment is actually giving mount.ntfs-3g all the trouble? i.e. which specific, exact files are mount.ntfs-3g using all that CPU power to read/write? The KDE System Monitor shows the mount command (/sbin/mount.ntfs-3g /dev/mapper/the_mount_point -o rw,big_writes,uid=1000,gid=1000,windows_names), but no further information. iotop shows extremely little disk activity overall, and only a few tens of K/sec for mount.ntfs-3g.
Any insight as to why it would take mount.ntfs-3g a quarter of my CPU power to run this webserver from ntfs, when it runs with 0% cpu usage on btrfs, would be greatly appreciated, as would any pointers on how I could debug & resolve the issue.
