I want to know how can I get native "Backup and Sync" experience (i.e. latency-free, always-sync) on my Ubuntu.
I have a repository that I am syncing with GDrive (pls don't shame, I know what I do is wrong).
I can use WSL+git on top of the GDrive folder very comfortably (ofc, we assume that only one side is being edited at all times, so there are no sync-collisions).
A simple git status takes 10 seconds on Ubuntu+rclone (on --vfs-cache-mode full).
I tried both
curl https://rclone.org/install.sh | sudo bash
rclone config
rclone ls gdrive:
rclone mount gdrive: /media/usr/gdrive
rclone mount --daemon gdrive: /media/usr/gdrive
rclone mount --daemon --attr-timeout 1m --vfs-cache-mode full gdrive: /media/usr/gdrive
If what I see is expected, at least let me know so I don't worry.
What is your rclone version (output from rclone version)
Using --vfs-cache-mode full is the closest you will get to it at the moment...
I tried this on a git repo I have on Google drive. The first git status took about 2 minutes and seemed to open and read every file in the checkout! A subsequent run took less than 1 second.
I tried git status consecutively multiple times, until I convinced myself that I will open a topic about it. git add was improbable that it would happen (.git/index.lock exists)
I know that it takes "10 seconds" because on the last (of n) invocation I did /usr/bin/time git status instead, and not because I counted it by a clock manually. I did it like 2-3 times, and then I lost interest - if something forces me to stop and busywait; basically I cannot make progress at all.
I know what I ask might be too time-consuming, no-benefit for you; however: Can there be way to contact to a rclone --daemon (or the non-daemon variant) that would show in an fancy-terminal, the total number of queue size, its progress, and the operations individually (files [local or remote path]/API URL (?)/size/percent-done/ETA)? Or something like progress -wM-equivalent (https://github.com/Xfennec/progress)?
I know for you a -vv output makes more sense, but for me it's a waste of deciphering time e.g. a list of 200 lines saying that (on an empty local directory) it would've created all the files, but it didn't do because --dry-run is active.
Instead, my proposal gives me much more input on "what is happening", even if I cannot do anything about the delays.