Files not being removed from cache?

What is the problem you are having with rclone?

Trying to cache an entire Google Shared Drive locally, while still allowing uploads/downloads to stay in sync. It looks like when things get updated on Drive, the VFS cache gets notified that it should be invalidated by nothing gets removed from local disk. Is this by design? Is there any way to change the behavior to have the local drive mirror the remote?

What is your rclone version (output from rclone version)

rclone v1.51.0

Which OS you are using and how many bits (eg Windows 7, 64 bit)

Amazon Linux 2 4.14.165-131.185.amzn2.x86_64

Which cloud storage system are you using? (eg Google Drive)

Google Drive

The command you were trying to run (eg rclone copy /tmp remote:tmp)

Currently using this command against a drive remote (NO cache remote wrapper in place):

/usr/bin/rclone mount gdrive:/music /mnt/gdrive \
    --attr-timeout 1000000h \
    --dir-cache-time 1000000h \
    --poll-interval 30s \
    --cache-dir /mnt/data \
    --vfs-cache-max-age 1000000h \
    --vfs-cache-mode full \
    --vfs-cache-max-size 240G \
    --log-level INFO \
    --log-file rclone-mount.log \
    --allow-other \
    --daemon \
    --rc \
    --rc-addr localhost:5572 \
    --rc-user test \
    --rc-pass test

A log from the command with the -vv flag (eg output from rclone -vv copy /tmp remote:tmp)

Can provide as requested, looking for guidance at this time. Just a general question really.

I also tried out the cache backend (wrapping the gdrive remote) and it didn't seem like it was caching the files locally in their entirety (like --vfs-cache-mode full does). Is that also by design?

Please provide the debug logs too.

By design (or maybe by oversight!). Stale cache files will remain until --vfs-cache-max-age passes or --vfs-cache-max-size is exceeded. The directory cache layer could be better integrated with the file cache layer - this is something I'm thinking about right now!

Thanks! If the cache storage size is reached, will it start purging invalidated objects or will it start purging oldest objects first?

Least recently used first.