What is the latest Plex-friendly mount and move parameters?

Off Topic
1

Hi,
I've been using these parameters since two or three years and I'm curious if anyone has found any improvements and are willing to share their setup?
I'm hosting my server at home on an Intel NUC with a 1Tb NVME-drive and utilizing the mover daily to move my stuff to Google Drive which is configured as a crypt towards a drive and not utilizing the team drive feature.

Mount:

rclone mount gcrypt: $LibraryRemote/ \
    --allow-other \
    --buffer-size 1G \
    --timeout 5s \
    --stats 1m \
    --drive-chunk-size 16M \
    --dir-cache-time 5m \
    --buffer-size 32M \
    --cache-db-purge \
    --vfs-read-chunk-size 128M \
    --vfs-read-chunk-size-limit 1G \
    --vfs-cache-max-age 1h \
    --vfs-cache-max-size 100G \
    --gid 100 \
    --uid 99 \
    --log-file=$ScriptsLogs/$LogName.log \
    --log-level NOTICE &

Move:

rclone move $LibraryLocal/ gcrypt: \
    --checkers 3 \
    --min-age 24h \
    --fast-list \
    --log-file $ScriptsLogs/$LogName-$(date ${LogDateFormat}).log -v \
    --tpslimit 3 \
    --transfers \
    --exclude-from $ScriptsRoot/settings/_excludes \
    --delete-empty-src-dirs \
1 Like
2

Honestly, for the reading/streaming part, a quality peering with google is more important than tweaking your settings imo, if you're using gdrive :smiley:

For the record, this is my plex mount command line (I know some stuff around needed/used in a mount) :

rclone.exe mount --cache-dir t:\mountcache --attr-timeout 8700h --use-mmap --stats 10s --buffer-size 32M --vfs-cache-poll-interval 30s --poll-interval 30s --dir-cache-time 8700h --vfs-cache-max-age 4h --async-read=true --drive-server-side-across-configs --vfs-cache-mode full --vfs-read-chunk-size 32M --vfs-read-chunk-size-limit 1G --local-no-check-updated --drive-chunk-size=64M --multi-thread-streams=4 --multi-thread-cutoff 250M --transfers 8 --drive-disable-http2=true --fast-list --timeout 30m --vfs-case-insensitive --vfs-cache-max-size 120G --log-level ERROR --drive-pacer-min-sleep 10ms --drive-pacer-burst 200 --log-file="c:\rclonelog\rclonelogplex.txt" --no-console --rc --rc-addr localhost:5573 "gdriveencryptedownapi:" P:  --network-mode

No problem, even for streaming uhd remux (I've a 2gb/sec down 600mb/sec up fiber connex.), and the scan part isn't killing my api quota (I'm on Windows btw).

2 Likes
3

I've read somewhere that the VFS flags shouldn't be used but I can't remember why. I still use them tho.
What does --no-console do for you and do you see any of my parameters that can be changed or removed into something more fitting? My bandwidth is significantly slower, at 250mbps.

4

hi,
the vfs flags are optional.
sometimes i use --vfs-cache-mode=off, sometimes i use --vfs-cache-mode=full
it depends on a lot of variable unique to your use-case.

if you want a set of well tested flags, copy them from
https://github.com/animosity22/homescripts/blob/master/systemd/rclone-drive.service

--buffer-size is listed twice, each with a different value.
fwiw, remove both of them

does nothing on a rclone mount

does nothing, is for the deprecated cache remote.

5

Thanks!
I used some of the parameters from the repo you linked to and forgot to change the cache dir, ending up in my /root dir taking up 35Gb and being unable to login to Unraid. Changed back to normal after remounting at more specific locations :smiley:

Just a quick question. Do I use the mover when utilizing the --cache-dir parameter? My library directory is configured like this:

  • Library
    • Local - Where my newly acquired stuff is located and later moved via RClone to my gcrypt:
    • Remote - The mount point for my RClone drive (:gcrypt)
    • Union - This is where my Local and Remote are merged with UnionFS and used in my Docker containers
    • Cache - This is new due to the change in the script. I can se two VFS-related directories in it right now.
6

you can use the mount to move files or you can use the mover script.
for me, i prefer to use a mover script

7

Ok, so in my directory setup I would place new files in my --cache-dir and RClone automatically takes care of moving files?
If I go back to using the mover script, I don't need the VFS parameters at all?
Thanks for your help! Greatly appreciated.

8

thanks,

should never copy files into the cache dir, rclone manages that.

instead, move the files to the mountpoint the P: drive.
rclone will upload the files.

correct.
the vfs file cache is optional; as some fellow rcloners do not use it, including myself on occasions.
for me, it depends on the quality of the internet connection.
with 1Gbps fiber + very low latency, no need for cache
with 5Mbps, unlimited hotspot from my cellphone + very high latency, i do use the vfs file cache.

1 Like
9

Wow, ok, that's very interesting. I'm not really happy I understood the meaning of the VFS parameters when GSuite is probably no longer an option. :rofl:

10

sorry, not sure exactly what you mean, did i miss a joke or what?

11

Sorry, I was being sarcastic. I've been using RClone for a long time and realized that the changes to GSuite might make my current set up unusable and I have to go back to local disks. EDIT: Meaning that I feel dumb for learning about the VFS parameters just now.

12

oh, ok, i see what you mean :wink:

13

Are you talking about the unlimited storage ? For now it's still on...