Unable to Optimize with plex on a rclone mounted cache

[Lono_Amstrad]

Unable to Optimize with plex on a rclone mounted cache, gdrive is only open on this machine,

Here's my rclone mount command for the cache:
/home/reach/bin/rclone mount pcache:/ /home/reach/remote/ --daemon --cache-writes --vfs-cache-mode writes --allow-other --dir-cache-time=12h --cache-workers=5 --buffer-size=15M --attr-timeout=1s --cache-tmp-wait-time 4h --umask 000 --cache-tmp-upload-path /home/reach/rclone_upload --log-level INFO --log-file=/home/reach/pdrive.log

settings from the rclone.conf for the endpoint (minus sensitive data) is:
chunk_size = 10M
info_age = 24h
chunk_total_size = 400G

log from plex:
Plex Media Server.1.log:Oct 15, 2019 03:20:34.354 [0x7f4262c0e700] ERROR - [Transcoder] [mp4 @ 0xe47f00] Unable to re-open /home/reach/remote/PMedia/Movies/Opt/Plex Versions/Custom_ Universal TV 117/.inProgress/Ramp (218).mp4.117 output file for the second pass (faststart)
Plex Media Server.1.log:Oct 15, 2019 03:20:34.356 [0x7f423affd700] ERROR - [Transcoder] Error writing trailer of /home/reach/remote/PMedia/Movies/Opt/Plex Versions/Custom_ Universal TV 117/.inProgress/Ramp (218).mp4.117: No such file or directory

log from rclone:
2019/10/15 06:19:40 INFO : Cleaned the cache: objects 21 (was 21), total size 6.944G (was 6.870G)
2019/10/15 06:20:40 INFO : PMedia/Movies/Opt/Plex Versions/Custom_ Universal TV 117/.inProgress/Ramp (218).mp4.117: Removed from cache
2019/10/15 06:20:40 INFO : Cleaned the cache: objects 20 (was 21), total size 7.012G (was 6.944G)
2019/10/15 06:21:40 INFO : PMedia/Movies/Opt/Plex Versions/Custom_ Universal TV 117/.inProgress/Ramp (218).mp4.117.36820.sidecar: Removed from cache
2019/10/15 06:21:40 INFO : PMedia/Movies/Opt/Plex Versions/Custom_ Universal TV 117/.inProgress/Ramp (218).mp4.117.36819.sidecar: Removed from cache

Any thoughts or tips?

[Animosity022]

Not sure we've ever gotten optimize to work out as it does some funky stuff along the way with the way it handles files.

Your best bet is to use something like mergerfs and setup a temporary local mount in place of the cache tmp upload and go that route.

My own service file looks like:

[root@gemini system]# cat gmedia.service
[Unit]
Description=gmedia mergerfs mount
Requires=rclone.service
After=rclone.service

[Service]
Type=forking
ExecStart=/usr/bin/mergerfs /local:/GD /gmedia -o rw,async_read=false,use_ino,allow_other,func.getattr=newest,category.action=all,category.create=ff
KillMode=process
Restart=on-failure

[Install]
WantedBy=multi-user.target
[root@gemini system]#

For my mergerfs setup.