Skip to content
View in the app

A better way to browse. Learn more.

Unraid

A full-screen app on your home screen with push notifications, badges and more.

To install this app on iOS and iPadOS
  1. Tap the Share icon in Safari
  2. Scroll the menu and tap Add to Home Screen.
  3. Tap Add in the top-right corner.
To install this app on Android
  1. Tap the 3-dot menu (⋮) in the top-right corner of the browser.
  2. Tap Add to Home screen or Install app.
  3. Confirm by tapping Install.

Plex Media Server

Featured Replies

Out of curiosity, did you run the command I posted to see if the share created by SNAP was in fact being created on the SSD or in ram? Unless it was a typo or wrong path on your part it could be a possible bug in SNAP.

 

Which command are you referring to that would verify the share was on SSD as opposed  to RAM???  I must have missed that..., please point me to it. 

 

Realize that all the checking you did via windows explorer just shows that files are being put into the share.  That is no indication where that share really resides.  I think I was possibly fooled by this line, "I periodically checked the plex folder on the SSD share and saw that it was steadily increasing in size" which I took to mean you were looking at the share via UnRaid gui (that is how I used to check for transcoding since my SSD is not exported to SMB) which in theory should only show shares on physically drives and not ramdisks.  But now I suspect you meant you were checking it via win explorer.  Is that correct?

 

Yes...  after starting PMS scan... I went back to Explorer and right clicked the properties tab of the "plex' folder inside share.  I saw that the folder was increasing in size.  I felt confident that share was on SSD because "/mnt/disk/PlexShare/" was the name I saw next to my SSD in the Disk Management tab in unRAID unMenu GUI. So I assumed the SSD was mounted as /mnt/disk/PlexShare.  When I pointed PMS library to the share, the "plex' folder was created inside the share

 

Another note tho', please excuse my inexperience with Linux..  I'm trying to run your script now!  I'm trying to find my SSD ID as you've instructed here:

 

" I mean open a console and go look at the contents of /dev/disk/by-id/, find your SSD and copy the entire string."

 

Which console are you referring to...?  I'm thinking you mean Telnet and then type in  "/dev/disk/by-id/" in order to fing my SSD ID? 

 

Also,  your script:

 

#Mount SSD as appdisk

mkdir /mnt/appdisk

mount -rw -o noatime,nodiratime -t reiserfs /dev/disk/by-id/[found in /dev/disk/by-id/]/mnt/appdisk

 

How and where do I apply...  In windows, I would copy to Notepad and then save in as .bat, how do I run this script in Linux?

 

......do you mind if we PM, I don't want to Hi-Jack this thread with my issues!

  • Replies 1.7k
  • Views 411.8k
  • Created
  • Last Reply

In a telnet session, type:

stat -f -c '%T' /path/to/share

 

It shows ramfs, its (obviously) in ram. If its on the SSD like it should be, it'll output reiserfs or whatever file system is on the SSD.

 

......do you mind if we PM, I don't want to Hi-Jack this thread with my issues!

I don't consider this issue solving off topic at all, you never know when someone else bumps into a similar situation. So please carry on the public channel ;)

 

I don't consider this issue solving off topic at all, you never know when someone else bumps into a similar situation. So please carry on the public channel ;)

 

Thanks brother!

 

In a telnet session, type:

stat -f -c '%T' /path/to/share

 

It shows ramfs, its (obviously) in ram. If its on the SSD like it should be, it'll output reiserfs or whatever file system is on the SSD.

 

Okay....I will try.  I disabled SNAP to begin the script provided by Jumperalex, but i will try this first to see if share was in the proper place..  StandBY

 

UPDATE:  ran command and the result was "reiserfs"  SO, share is on SSD!

Yeah get an answer to Inluencer's query first and then we can talk about how I mounted my drive.

Saw your update ... well great googly moogli  :o

 

Ok so taken as a given that you are indeed pointed at a physical drive with a partition.  Can you confirm the size of that partition?  sorry my command-line-foo is actually not great and I don't have access to my terminal to do some just-in-time learning ... maybe Influencer can help out.

 

 

Saw your update ... well great googly moogli  :o

 

Ok so taken as a given that you are indeed pointed at a physical drive with a partition.  Can you confirm the size of that partition?  sorry my command-line-foo is actually not great and I don't have access to my terminal to do some just-in-time learning ... maybe Influencer can help out.

 

I ran "df -h" from telnet session and got this:

 

root@Tower:~# df -h

Filesystem            Size  Used Avail Use% Mounted on

tmpfs                128M 1020K  128M  1% /var/log

/dev/sda              7.3G  105M  7.2G  2% /boot

/dev/md1              2.8T  1.4T  1.4T  50% /mnt/disk1

/dev/md2              1.9T  1.4T  489G  74% /mnt/disk2

/dev/md3              2.8T  1.1T  1.7T  38% /mnt/disk3

/dev/md4              1.9T  1.4T  465G  76% /mnt/disk4

/dev/md5              1.9T  1.3T  577G  70% /mnt/disk5

/dev/md6              1.9T  468G  1.4T  26% /mnt/disk6

/dev/md7              1.9T  59G  1.8T  4% /mnt/disk7

/dev/md8              1.9T  11G  1.9T  1% /mnt/disk8

/dev/md9              2.8T  821G  2.0T  30% /mnt/disk9

/dev/md10            2.8T  129G  2.7T  5% /mnt/disk10

/dev/sdd1            233G  3.8G  230G  2% /mnt/cache

shfs                  22T  7.8T  15T  36% /mnt/user0

shfs                  23T  7.8T  15T  36% /mnt/user

/dev/sdh1              56G  68M  56G  1% /mnt/disk/AppShare

 

I remounted SSD as AppShare in SNAP 

 

My SSD is /dev/sdh1

 

I wonder why  my parity drive isn't listed

...I wonder why  my parity drive isn't listed

Parity drive is not a mounted filesystem. In fact, it has no filesystem.

As another troubleshooting step, point plex to that now known good share and let plex start its metadata gathering. Then continue to use df to monitor usage so we can see if the share is indeed growing in size.

 

Also open another telnet window and run this command:

free -ms 30

 

Leave that running and it will update the memory usage every 30 seconds. This way we can see if 1) the SSD is indeed growing in size and 2) If memory usage is growing to a point that OOM kicks in.

 

If your system goes down again hopefully this will give us some insight as to what is happening.

As another troubleshooting step, point plex to that now known good share and let plex start its metadata gathering. Then continue to use df to monitor usage so we can see if the share is indeed growing in size.

 

Also open another telnet window and run this command:

free -ms 30

 

Leave that running and it will update the memory usage every 30 seconds. This way we can see if 1) the SSD is indeed growing in size and 2) If memory usage is growing to a point that OOM kicks in.

 

If your system goes down again hopefully this will give us some insight as to what is happening.

 

Will do...  Question about  memory updates..  Do I need to stop the 30sec memory updates after test?  If so, what command do i use to stop that process. 

 

I will perform the test in a few hours so please stand by.  Studying the the 70-640 exam I have to take next month! ;)

CTRL+C to stop the command from running. And for future reference its the same with any command you run in linux, if it doesn't terminate itself you can press CTRL+C to stop it.

Guys..  I'm 10min into scan and all-is-well  SO FAR!!  Have 3 telnet sessions open monitoring:  df -h to monitor SSD size increase, free ms -30 to reset memory q30 sec, and TOP to monitor CPU and Memory%.

 

Kinda feel stupid, why didn't this work before!?  I did the same things I'm doing now as i described in earlier posts with the exception of 30sec memory resets. As of now TOP has memory% @about 4.5% average.  CPU% fluctuate between 10-75%.  These Values are "Plex Media Scan" values by the way!!  SSD share size is increasing with every check as well.  I'll update this post as progress continues!

 

Cache drive finally spun down after a full 37 hour grind! 

It was probably a race between plex/snap that caused plex to install and run before snap installed and ran. Once you disabled it and then set it back up everything mounted how it should and is now working.

 

Once things keep going, I'd say tomorrow if you don't have issues before then, reboot the server and see what happens, check the SSD share and your mem usage to see if everything is working correctly. Without testing everything myself (and not remembering the exact boot sequence), the plex package MIGHT get installed and started before the snap plug-in. This would cause the same issue you had before.

Without testing everything myself (and not remembering the exact boot sequence), the plex package MIGHT get installed and started before the snap plug-in. This would cause the same issue you had before.

 

DAMN... 

 

Well......... I'm thinking with the "Glass 1/2 full" bro!! :)  I will restart unRAID array with confidence that SNAP wont fail me and mount the drive on-time!  I got this far...and I have such a large Kool-AID grin which I want to keep after reboot. 

 

Have I been the only one to report this problem with PLEX??  I did read about member reports with other plugins causing OOM issues when share was outside the array, but not PMS.  Am I the redheaded stepkid or what?? 

I think with the "glass 1/2 full" too. I just like to remember the glass can also be empty.

 

As Leela from Futurama said, "It worked! Gravity normal. Air pressure returning. Terror replaced by cautious optimism!"

Well Guys....  I've got some good news & I've got some bad news!

 

Good news 1st:

 

I was able to mount the SSD using SNAP and unRAID was able to recognize the share.  I was able to add my media and run the Plex Scanner without OOM issue.

 

Bad News: 

 

After scan completes, I restart unRAID, check SNAP(Green indicators say drive is Mounted and Shared), Enter share UPN name in Windows run box(Share was visible and accessible), BUT, open Telnet and run "df -h" and share does not show up as mounted.  So, I go to the Disk Management tab in unMENU and see that the SSD is not mounted.  So I mount the SSD  as writable and run df command again and share shows up as mounted.  So, figuring that when I start Plex web GUI the library would still be there, but the EULA shows up and I have to start Plex library from scratch.  I don't understand this one because df info says SSD size is the same prior to restart and Plex plugin is still pointed to the same share?  I tried to stop/start plugin but I got the same result, library is empty! 

 

....sorry guys! :-\

 

 

A couple questions for clarification. When you mounted the drive in unmenu did you disable the drive in snap or mount it to a different share name?

 

Id have to look at how snap checks if a drive is mounted but if it only looks for the share name to validate then it may get false positives in a case like this.

 

I did check and packages are installed before plugins, as well as plex being alphabetically before snap it would cause this.

 

Maybe we should appeal to the snap dev to change the name to 'asnap' or something to make it start before other plugins as this would probably be a typical use for snap.

[cough] or do it all via two lines in the go-script :) [cough]

A couple questions for clarification. When you mounted the drive in unmenu did you disable the drive in snap or mount it to a different share name?

 

Id have to look at how snap checks if a drive is mounted but if it only looks for the share name to validate then it may get false positives in a case like this.

 

No....  I did not unmount/unshare the drive in SNAP nor did I modify the share name in SNAP or disk management.  The share name I created in SNAP(AppShare) shows in DM and df -h as /mnt/disk/AppShare after I click the mount button.  I checked DM to see if unRAID had recognized that SNAP had mounted the drive, I then mounted when I saw that it hadn't.  df -h had pretty much let me know initially, I just went to disk management for confirmation.  BUT, I mounted the drive in disk management also to see if I could get PMS to recognize the library but it didn't, even tho' df -f results said the drive/share was mounted.  The PMS library info is still on the SSD. 

 

I hope this answers your question properly ;)   

 

I did check and packages are installed before plugins, as well as plex being alphabetically before snap it would cause this.

 

Maybe we should appeal to the snap dev to change the name to 'asnap' or something to make it start before other plugins as this would probably be a typical use for snap.

 

I'm no expert of course, but It seems as though SNAP loaded the SSD just fine.  My perception is that there is a communication barrier between SNAP & unRAID.  Remember, the share was visible and accessible after reboot.  Unless, SNAP just mounted the share and not the SSD which is why df -h results were empty?  or am I just waaaayyy out confused?? ??? 

[cough] or do it all via two lines in the go-script :) [cough]

I would like to try your option next! 

What I think happened this last reboot is plex started and created the directory. If it doesn't exist when plex starts it will be created. Snap most likely checks the path to verify if its mounted, so if plex created the directory snap would think its mounted, giving you the green light.

 

I'm a bit confused on what you did in disk management as if the directory exists and is not empty the mount will not be successful.

 

In the end I'm pretty sure snap sees the directory and thinks it is mounted properly when in reality plex created the directory (thus creating the share). Smb doesn't know the difference from a directory created in ram vs a mounted drive, it will export the share as long as it exists.

 

I think what got you mixed up in that last round of troubleshooting is you tried to remount to a share that already existed when you should have disabled snap and then mounted in disk management, or just disabled snap, removed the directory and enabled snap again.

 

I agree that just doing it from the go file will work the easiest, I'm just trying to get down to the cause to see if its a bug or just poor timing with all of the add-ons initializing. Limetech has stated in the past they want to move away from the go file being used for extras like this, of course its your system and as long as the option is there use it. If limetech decides in the future to use a rc script to start emhttp (the only thing limetech officially uses the go file for), the option of using the go file to mount the drive might be out of the question.

 

If you want to test things further pm me, but I wouldn't blame you for going the script route. I'm going to scrounge up a USB drive to test with myself.

You might need to scrounge up an HDD to test since SNAP treats USB and SATA drives differently from what I read.  But I could be wrong.

From what I can tell the only difference is a USB drive is automatically shared, sata drives must be manually shared. Even if its not shared it should be accessible to the system, so for testing purposes should do fine. At any rate, my server is out of sata ports so it'll have to do!

I am using Sony Smp-n200 to play the movies. In the PMS there is a DLNA profile for it. I want to delete it & use PMS without transcoding. How can I do that????

You'll probably have to write a custom profile to direct play only. I can look into it further once I get home unless someone else can explain further before then or you figure it out.

Archived

This topic is now archived and is closed to further replies.

Account

Navigation

Search

Search

Configure browser push notifications

Chrome (Android)
  1. Tap the lock icon next to the address bar.
  2. Tap Permissions → Notifications.
  3. Adjust your preference.
Chrome (Desktop)
  1. Click the padlock icon in the address bar.
  2. Select Site settings.
  3. Find Notifications and adjust your preference.