Server Stats (Current CPU/Memory/etc usage)

JonnyWong16 asked for this feature over 5 years ago — 19 comments

JonnyWong16 commented over 5 years ago Admin

I'm pretty sure I read this request somewhere else, but couldn't find it.

So I ran across this package and was wondering if a dev could maybe glance at it and see if this is something we could use to add a server stats page? Just a thought, would be really cool to have.


Reference previouse request on GitHub: #175

| 12nick12 commented over 5 years ago

It would also be great if there was someway to throw this information into a graph. It would be great if yo could throw the CPU usage on the Daily Stream type breakdown graph, just to see how much transcoding is pegging your CPU.

| Chrisophogus commented over 5 years ago

As part of this would it be possible to get bandwidth stats for the Plex service?

YipYup commented over 5 years ago

If this gets implemented, I would like to request it be something that can be optionally turned on or off. I use separate tools to monitor and log performance and would not want to log it twice.

| Foebik commented over 5 years ago

@JonnyWong16 Yep, I opened a similar request back last Sept.


| Foebik commented over 5 years ago

NVM, I see you found that one too.

| Foebik commented over 5 years ago

I did however get the HTPC-Manager working. Not 100% the way I wanted to consume the data, but better than nothing. http://i1285.photobucket.com/albums/a583/Foebik1123/Simple_zpsbrvj6fhv.jpg http://i1285.photobucket.com/albums/a583/Foebik1123/Expanded_zpssnqyl83l.jpg

deepwather commented over 5 years ago

You guys can use my projetc, and intigrate it to your plexpy.. If you want, its already the same design & it's fully working! https://github.com/deepwather/plexDash


| Foebik commented over 5 years ago

Your project looks pretty nice. I ran across it the other day and then was sad when it said "Linux only".

ColdIce1 commented over 5 years ago

Just gonna copy + paste my answer from the issue to this tracker:

This is very much interesting, but people might run Plex on docker or similar which might create issues.

I suggest closing this issue, as PlexPy is focusing on improving the Plex administrator experience and NOT server monitoring or real time data regarding CPU, memory, IO statistics and so on.

If you really are interesting in such things, look into Collectd and set it up with Graphite. Then you got your logging going.

| PotatoQuality commented over 5 years ago

Well, i;m do not agree with Coldice1. Plexpy is a application for monitoring, analytics and notifications for Plex Media Server. Its not ore problem you have issues with plex with docker. a lot of people use this application without a docker.

| basmith7 commented about 5 years ago

The problems I see with this are that PlexPy has to be running on the same computer as the Plex Server, and its going to be very platform dependent (Linux, Mac, Windows) code.

| spot88 commented about 5 years ago

This would be awesome, and for you people that don't think this is needed or is using something else. Don't use this feature then. I'm using HTPC now, and would love to just use one interface to check everything. I agree this should be able to turn on\off.

evanrich commented about 5 years ago

alternatively, you guys can always use collectd + influxdb + grafana to monitor your box

tripflex commented almost 5 years ago

I agree server stats can be monitored many other ways, I personally use Observium for monitoring most of my hardware (mainly network but works well with servers too). With that said, what about something like buffering stats? I think that would be huge :D

Added a feature request specifically for buffering stats: http://feathub.com/JonnyWong16/plexpy/+157

| Altheran88 commented over 3 years ago

Would be nice to hace in Tautulli a live dashboard about, CPU usage (1 / 5 / 15m) https://en.wikipedia.org/wiki/Load_(computing) RAM Netkork usage Then Daily graphs on Max (15min CPU) Max Ram Max(15 min Network Outbound)

A lot of how to's could be deviced trom https://atlanta.my-netdata.io/default.html#menu_system_submenu_cpu;theme=slate https://github.com/firehol/netdata

breadlysm commented over 3 years ago

As others have said, a feature like this may be very machine dependent. Additionally, I see a lot of bug reports from it.

Alternatively, the homepage panel layout is setup perfectly for something like graphana embedding. I picture a custom HTML panel option that could be used for any purpose. Graphana allows you to embed graphs on other sites. Putting the iframe code in a custom HTML panel would look great (in my head at least(and likely would not look great on mobile)). http://docs.grafana.org/reference/sharing/#embed-panel

Not a direct solution for this feature but, can indirectly fulfill it. A custom panel could be used for lots of things like Plex news and other feeds, sports scores, youtube, google docs including forms, chat rooms.

melophat commented about 3 years ago

@basmith27 i know this is a reply to a 2 year old post, and maybe your post was accurate at the time of posting. However, PlexPy/Tautulli does NOT need to be running on the same server as the plex software, it only needs access to the logs directory. I've currently got Tautulli and other plex support software/automation software running on a separate virtual machine than my plex server, and by sharing the plex logs directory and mounting it on the Tautulli machine, I was able to get Tautulli working correctly. there's an option in Tautulli to set the plex logs directory that makes it all easy.

hobesman commented over 2 years ago

@melophat I think his point was that Tautulli could only tell you the CPU, RAM, etc. of the machine Tautulli is running on. If that isn't the same machine as the Plex server, then gathering CPU and RAM info about the machine Tautulli is on is of no relevance to your Plex server. That's what he was trying to point out. If you start his post with this it makes more sense: "In order to report CPU usage of your plex server, plexpy has to be running on the same computer..."

Join the discussion!

with GitHub to comment