Hi there,
I am really hoping someone can help me troubleshoot why my Plex performance is so poor when resuming / seeking during Movie playback.
I am running Plex on a VPS server, which is using RClone VFS to stream from an encrypted GDrive folder.
I have seen this on an MP4 and MKV file recently, but may affect other file types (I can test, if it’s useful). I am fairly certain this issue is new.
Essentially, if I start playback from the beginning of a movie, it takes a couple of minutes to start. This has always been the case even though I have Plex set to Direct Play and “Play Original Quality” (rather than transcoding) and I assume this is because of the amount of time it takes for my server to download the data from the encrypted Gdrive, fill the Plex buffer and then send the data out to my client (which is geographically some way from where my Plex server is). Once playback starts, it is flawless for the duration of the movie.
HOWEVER, if I pause/stop the movie and then want to come back and finish it later (ie: select RESUME from HH:MM:SS in Plex), playback never starts. I get the Plex spinning wheel for an indefinite period. I have Netdata running on my VPS so I can see that server resource usage (notably, network bandwidth) is very low while this is happening (whereas, I would have expected to see the network usage showing loads of data being downloaded from the encrypted Gdrive as Plex tries to find the correct place to start playback from).
This issue is present on multiple clients (Panasonic TV aka FirefoxOS, Amazon Fire Stick 4K, iPad running latest iOS Plex client).
I’ve included some info on my setup below. I considered turning on DEBUG logging for the Rclone Service but I thought I’d get some initial feedback first because I wouldn’t know how to run a useful test at the moment. Similar for Plex - I don’t know how to turn on debug logging (although I’ll look it up if necessary) but I need a steer in the right direction of what to test first.
I’ve tried tweaking the buffer size and VFS read chunk size but until I have some idea of how to debug and interpret where the bottleneck is, I think I’m wasting my time…the changes I made seem to have had no useful effect.
I would appreciate any help from Rclone (or Plex) experts on where I start looking for what the problem is!
Thank you very much in advance for your time.
Rclone Version: 1.45 linux/amd64
Plex Version: 1.14.1.5488
OS: Ubuntu 18.04.2 LTS (VPS)
Rclone Service used for Gdrive Encrypted Media Folder:
[Unit]
Description=Mount ENCRYPTED Gdrive Media Folder (with RClone VFS and No RClone Cache) to /mnt/GdriveMedia
After=syslog.target local-fs.target network.target gdrive-vfs.service
[Service]
Type=simple
User=root
ExecStartPre=-/bin/mkdir -p /mnt/GdriveMedia
ExecStart=/usr/sbin/rclone mount
–config /home/plexuser/.config/rclone/rclone.conf
–allow-other
–vfs-cache-mode writes
–vfs-read-chunk-size 256M
–vfs-read-chunk-size-limit off
–buffer-size 2G
–dir-cache-time 48h
–drive-chunk-size 32M
–log-level INFO
–log-file /home/plexuser/gdrivecrypt-vfs.log
–timeout 1h
–rc
GdriveCrypt: /mnt/GdriveMedia
ExecStop=/bin/fusermount -u -z //mnt/GdriveMedia
ExecStop=/bin/rmdir /mnt/GdriveMedia
Restart=always
[Install]
WantedBy=multi-user.target