Interesting… Mine is very stubborn too currently (also with SSD Nodes). Started about a week ago, lots of buffering and slow starts. I thought it had something to do with Google or Rclone because a friend with a dedicated server (so obviously not the same provider) experienced the same.
The first month mine used a lot of bandwidth as well but after all the deep scanning finally finished, data usage seems rather normal and now towards the end of my second term I’m not even at 50%.
@Qlogic said:
it was working fine for about a month and how it can’t do anything once the transcoding starts, it’s good for 3-5 min and then its like all the CPU power disappears and it can’t transcode anymore. only thing that works is if you are direct playing and there is no CPU usage.
Mine is in Dallas, TX and it’s doing the exact same thing as yours. I even tried it with a local video to rule out rclone and Google drive and it still happend.
It’s like right after there trial they throttle the vps very annoying. I have moved back to my local machine
I’m using SSD Nodes as well as another provider in which I have the entire server (not shared) and full control over the CPU/memory/bandwidth. That server, as well, has been having issues the past week or two. Before jumping to say that SSD is throttling (which I’m not saying they’re not) I think there is a correlation to the latest Plex release (PMS, not Plex Cloud) and my problems. It also could be something Google is throttling. Plex Cloud for me is pretty reliable and I have had better success with that than with the VPS on numerous occasions. During the last past week, Cloud also has been having issues with the Google library.
Another thing to consider is that SSD Nodes is a shared environment. You don’t have the entire server to yourself. So, someone else on the same server running something CPU intensive could cause you to see a performance decrease (which you could also be causing for them when you stream).
@kelinger said:
I’m using SSD Nodes as well as another provider in which I have the entire server (not shared) and full control over the CPU/memory/bandwidth. That server, as well, has been having issues the past week or two. Before jumping to say that SSD is throttling (which I’m not saying they’re not) I think there is a correlation to the latest Plex release (PMS, not Plex
I use SSD Nodes in Seattle, and it has been okay. (It’s the Google 24-hour bans that have me ripping out my hair!!!) I haven’t been throttled by them in several weeks, but when I did, they sent me an email saying that my server was causing performance issues for other clients. It hasn’t happened since I turned off deep analysis and limited my library maintenance to a few hours in the early morning. Also, they completed a hardware upgrade/maintenance task in the Seattle data center about 3 weeks ago which seems to have helped.
I just watched a TV show and attempted from three different servers. A high-end VPS (full server owned by me) in Europe mounting an encrypted Google drive with rclone; SSD Nodes in Dallas using a non-encrypted Google drive via rclone; Plex Cloud using a non-encrypted Google drive.
All 3 had issues playing (where as episodes of the same show in presumably the same format and relative size a few weeks ago streamed uninterrupted from all 3 servers). The only one that allowed me to watch the entire episode was Plex Cloud and this still had long “rebuffering” delays (1-3 minutes) at approximately 4 different points in the episode. The Plex engine on the VPSs was version 1.6.1. I believe I’ve had this issue since 1.6.0. Plex Cloud, well, I’m not sure how to read the version number. Its been the same since before it left “beta” but there has been work done on the back end and I would assume (though we know what that means) that code shared between both PMS and Plex Cloud could very well have been introduced to the backend at about the same time.
My point is that if SSD is throttling and sending out email when they do, I don’t think that they’d also throttle and NOT send out an email. I wish I could find an older Plex to reinstall and verify once and for all.
When i look into the PMS logs from my VPS there are unbelievable much Error Messages in there… Any idea why?
The Media analysis really eats into your bandwidth. I like the idea of limiting the schedule massively. At least that way you get the media analysis but without losing 8TB bandwidth in a matter of days.
has anyone else made the switch to using PLEXDRIVE to mount Google Drive?
Yes, for the last couple days. Just upgraded to 2.0 about 5 minutes ago and about to give it a spin!
Very impressed with it so far. I’ll be trying it with my encrypted media set soon.
I’ve been testing it out for a couple of days now. It seems to be working well, especially with build v2 which was released less than 24 hours ago. However, I managed to still get myself banned by Google this afternoon. I don’t know what caused it exactly since I still have Couchpotato and Sonarr using the rclone mounts (plexdrive is read-only for the time being). I was also testing a bunch of movies and rescanning the library frequently using the plexdrive mount, too. Until the ban happened, it was working really well! Movies were starting up in 2-3 seconds, just like rclone. I didn’t watch anything for longer than 10 minutes, but playback was great. Jumping ahead seemed pretty good too, usually taking ~10 seconds, but I did have some that took as long as 30-40 seconds.
I’m going to back off on the Library scans again and go back to my script that only scans individual directories that have changes. The Plex option for detecting changes and doing partial scans of the library doesn’t seem to work (I’m using rclone crypt on top of plexdrive; not sure which one or both causes that feature to not work). I still need a way to delay the scanning until plexdrive updates its cache though… I’ve hacked it with a sleep delay for now…
@djsecrist said:
I’ve been testing it out for a couple of days now. It seems to be working well, especially with build v2 which was released less than 24 hours ago. However, I managed to still get myself banned by Google this afternoon. I don’t know what caused it exactly since I still have Couchpotato and Sonarr using the rclone mounts (plexdrive is read-only for the time being). I was also testing a bunch of movies and rescanning the library frequently using the plexdrive mount, too. Until the ban happened, it was working really well! Movies were starting up in 2-3 seconds, just like rclone. I didn’t watch anything for longer than 10 minutes, but playback was great. Jumping ahead seemed pretty good too, usually taking ~10 seconds, but I did have some that took as long as 30-40 seconds.I’m going to back off on the Library scans again and go back to my script that only scans individual directories that have changes. The Plex option for detecting changes and doing partial scans of the library doesn’t seem to work (I’m using rclone crypt on top of plexdrive; not sure which one or both causes that feature to not work). I still need a way to delay the scanning until plexdrive updates its cache though… I’ve hacked it with a sleep delay for now…
To allow Plex to see the changes you would need to run something like unionfs-fuse
That’s what I’m doing currently and plex will now automatically scan new items added to the union of local & plexdrive
Hello App using an VPS with an rclone mount… what you are have in Download speed when you are use the Download button in PlexWeb?
@Massaguana said:
Hello App using an VPS with an rclone mount… what you are have in Download speed when you are use the Download button in PlexWeb?
For me, I generally get about 25MB/s to 28MB/s from my remote server to my home.
My home download speed is 300Mbps
But this isn’t with an rclone mount, it with either ocaml or plexdrive as the mounting software connected to google drive
with my rclone mount i have only 500 KB´s… my home speed are 450 Mbit don´t know whats the problem
@Massaguana said:
with my rclone mount i have only 500 KB´s… my home speed are 450 Mbit don´t know whats the problem
Peering issue between you and your remote server?
What type of speeds do you get when you download direct from your cloud provider (ACD or GDrive)?
Have you tried any of the peering workarounds, using CloudFlare or creating an SSH Tunnel using aVPS that is closer to you?
CloudFlare Solution: https://quickbox.io/wiki/set-up-a-cdn-for-plex-with-cloudflare-nginx/
Plex Routing thru an SSH Tunnel: Routing plex traffic through an SSH tunnel · GitHub
Mayby… iI am currently busy with the server change to another provider. From Gsuite i have in moment 21 MB/s direct Download Speed.
Is it possible that the problems because my GSuite are from US but i am in Europe?
@hthighway I have been playing with plexdrive and it seems pretty good, at least for playback.
I had been scanning a different mount and then switching to a plexdrive mount for playback.
I have now taken a gamble and am running a scan on plexdrive, so now waiting to see if I get hit by google’s ban hammer
Certainly the api usage is very low, its just the way google see’s the plex scanner as a lot of downloads that could cause issues. If anything, I think this is what could cause the ban hammer to fall.
So we will see what happens…
Seeing some bad throttling with SSD Nodes lately, though does anyone know what CPU OVH uses for their VPS?
I’m seeing that SSD Nodes uses a 4 core Intel Xeon CPU E5-2630 v3, though getting it throttled down to 240 MHz per core.
Anyone have any issues with OVH and having 1080p content transcoded?
How does Kimsufi’s “VPS SSD 3” plan compare with OVH’s?
