20 replies to this topic

[Mies]

I have a problem with the share (NASMedia) to my synology nas.

Every 5 minutes, more or less, it is going to search the LAN.

The wheel in the upper-left corner is running.

After 30 seconds, more or less, it is back again and in that time the box is unresponsive.

 

I have tried all possible solutions that, I could find on the internet.
fstab, smb share ,nfs share,, different options in the share, put in a ssd hdd, you name it I have done it.

 

Then I was thinking, what if the amount of files, (movies) is the problem of the LAN check.

When there are a lot of files, the box want to check that there are changes and is searching through all the files.

So I created a new share (NASOpenPLI) with no recorded movies at all. I disabled the other share in the mounting points.
Now it seems to working correctly, I had no more an unresponsive box. (for one evening, so far)

 

What is my setup:

Vu+Solo2
Vu+uno4k

Synology NAS 1817+

Openpli latest version

 

Automount:

 <mount>

  <active>False</active>

  <hdd_replacement>False</hdd_replacement>

  <host>nas</host>

  <ip>192.168.xx.221</ip>

  <sharename>NASMedia</sharename>

  <sharedir>volume3/Media</sharedir>

  <options>rw,udp,nolock,intr,timeo=10,retrans=20,rsize=8192,wsize=8192,nfsvers=3</options>

 </mount>

</nfs>

<nfs>

 <mount>

  <active>True</active>

  <hdd_replacement>False</hdd_replacement>

  <host>nas</host>

  <ip>192.168.xx.221</ip>

  <sharename>NASOpenPLI</sharename>

  <sharedir>volume3/OpenPLI</sharedir>

  <options>rw,udp,nolock,intr,timeo=10,retrans=20,rsize=8192,wsize=8192,nfsvers=3</options>

 </mount>

</nfs>

 

I have a total of 4.32 TB of recorded movies
Total files: 3726 files.
Movies, 6 files per movie= 621 recorded movies.

I have 6 folders in the share

AutoDocs
AutoTimer
User1

User2

User3

User4
 

I have created all marking point to all folders above.

The autotimer is the default recorded folder.
The amount of movies, in AutoTimer, is more or less 10 movies because it gets cleaned every day.

What could be the problem ? 

 

Regards,
Mies

Edited by Mies, 31 August 2024 - 22:16.

[40H3X]

I have no idea why the mounts are unstable, is it UTP or WiFi? The file you are referring to is missing a <nfs> after the first <mount>, but I think it is a typo and I do not know why you edited out the 3 octet with xx, as this ip is a RFC 1918 and can be anywhere.
 

/etc/enigma2/automounts.xml

<?xml version="1.0" ?>
<mountmanager>
<nfs>
 <mount>
  <active>True</active>
  <hdd_replacement>False</hdd_replacement>
  <ip>192.168.1.250</ip>
  <sharename>Recordings</sharename>
  <sharedir>volume2/video/Recordings</sharedir>
  <options>rw,nolock,soft</options>
 </mount>
</nfs>
</mountmanager>

If I may comment on your automounts.xml file, why so many options, do you really need them to be determent by the client?  When you not set them the NAS may decide what they will be. For instance I would let the NAS decide (server side) what the version NFS protocol is used. So below is the result of the automounts rule above, with only 3 options (default).

 

root@vuduo4k:/etc/enigma2# mount | grep nfs
nfsd on /proc/fs/nfsd type nfsd (rw,relatime)
1192.168.1.250:/volume2/video/Recordings on /media/net/Recordings type nfs4 (rw,noatime,vers=4.1,rsize=131072,wsize=131072,namlen=255,soft,proto=tcp,port=0,timeo=600,retrans=2,sec=sys,clientaddr=192.168.5.64,local_lock=none,addr=192.168.1.250)
root@vuduo4k:/etc/enigma2#

and my NFS mount is around 3.4TB at the moment of writing

root@vuduo4k:~# df -Th | grep Recordings
Filesystem           Type            Size      Used Available Use% Mounted on

192.168.1.250:/volume2/video/Recordings
                     nfs4           10.5T      3.4T      7.1T  32% /media/net/Recordings
root@vuduo4k:~#

Maybe start a debug and wait until the spinner comes along and post the log here, how to is here https://wiki.openpli...ebugging_Enigma

 

[WanWizard]

I also have no idea what is searching the LAN when, and where. Installed a 3rd-party plugin? Have autotimer search every 5 minutes?

 

I have about 20TB in files on my NAS, and I see only a spinner when I open a folder with lots of files in the movie list.

 

Which happens because it needs to enumerate them all, and, by default, it will also read them to calculate the length of each (this can be switched off in the settings). This can take a while with lots of files, especially over wifi, with low(er) bandwidth and high latency.

[tonskidutch]

@Mies

have you tried

 

<options>rw,udp,nolock,intr,timeo=10,retrans=20,rsize=8192,wsize=8192,nfsvers=3</options>

changing it into

<options>rw,nfsvers=3</options>

 

because the real version 3 or 4 NFS server doesn't need the chabang of options...  it negotiates the values between client and server.

I even have checked it and my QNAP is "hard" mounted altough at sometime i used the option >soft< . It might be firmware issue of your Synology.

 

What do yu mean with this

 

The autotimer is the default recorded folder.

How can that be when you use

<hdd_replacement>False</hdd_replacement>

?

[Tech]

That is depending on the chosen folder for your default recordings, it is not dependent on a harddisk.

[tonskidutch]

Well to be honest I believe the symbolic links to those various mount paths are faulty.

And if they are indeed mounted or not should be shown by a simple telnet command.

[WanWizard]

There are no "symbolic links to various mounts".

 

There is one, /media/hdd, optionally, if you set hdd_replacement to True and there is no local disk present.

 

And symbolic links have zero impact on disk performance, so that can't possibily explain the behaviour.  With broken symlinks, it either always works, or it never works.

 

Getting a spinner every 5 minutes is the result of an enigma process accessing the disk and taking to long to finish.

 

I wouldn't be suprised that autotimer is configured wrong, as that will scan every eit file then searching for new timers to create. Which is pointless, as EPG doesn't change every 5 minutes.

[Mies]

I also have no idea what is searching the LAN when, and where. Installed a 3rd-party plugin? Have autotimer search every 5 minutes?

 

I have about 20TB in files on my NAS, and I see only a spinner when I open a folder with lots of files in the movie list.

 

Which happens because it needs to enumerate them all, and, by default, it will also read them to calculate the length of each (this can be switched off in the settings). This can take a while with lots of files, especially over wifi, with low(er) bandwidth and high latency.

[Tech]

Well to be honest I believe the symbolic links to those various mount paths are faulty.

And if they are indeed mounted or not should be shown by a simple telnet command.

In any case, I can tell you that symlinks do not pose any problems as far as HDD replacement is concerned.

I have been working this way for years in combination with a Synology NAS.

[Mies]

Thanks for your input so far.

Because of the input, of all above I looked again, I changed my share.

 

At first, the mountpoint was: volume3/Media/

I created markers to go easy to my TV recording.

But the Media, which is the top-level of the share, has a lot more folders with other media, non enigma recordings, in it.

 

So I changed the share to go directly to TV recordings, so it cannot find the other media anymore.
For now, the spinners seems to stay away.

I don't use WiFi and have low latency.
The EPG is set for 6 hours.

 

I will see if the problem stays away.

[WanWizard]

That removed the symptom, but not the cause. You still haven't found what is accessing your mount every 5 minutes.

[tonskidutch]

 I changed my share.

 

At first, the mountpoint was: volume3/Media/

I created markers to go easy to my TV recording.

But the Media, which is the top-level of the share, has a lot more folders with other media, non enigma recordings, in it.

 

So I changed the share to go directly to TV recordings, so it cannot find the other media anymore.

Well we would like to know and are interested how you achived those markers which are changed to the previous setup of that share.

[Mies]

This is what I did:

 

previous mount point.

volume3/Media/

volume3/Media/TV-Opname

volume3/Media/Video

volume3/Media/Music

volume3/Media/Books

volume3/Media/Kodi

volume3/Media/Backup

 

Now it looks in all folders for media, also the video section.

 

New point is: volume3/Media/TV-Opname

it only searches in TV-Opname, which only consists of engima recordings.

 

Then I adjusted / added the bookmarks for the recording paths ( I had to change language to get the english word)

With the yellow button everybody can easy change to there location.

 / media / hdd / movie

 / media / net / NASMedia / AutoDocs /

 / media / net / NASMedia / AutoTimer /

 / media / net / NASMedia / User1 /

 / media / net / NASMedia / User2 /

 / media / net / NASMedia / User3 /

 / media / usb /

My standard recording path, for all options, is now: /media/net/NASMedia/AutoTimer/
Also I have a physical disk of 500GB SSD, for the EPG, which is 6.627 MB small.

 

So change is small, just less folders to search for.
But we have to wait if it is stable and that I don't make the wrong assumions I can easily mis other parameters .

[WanWizard]

"it looks", "it searches"...

 

WHAT is "it" ? There is nothing in OpenPLi that accesses disks every 5 minutes, let alone scan or read ALL files on that disk. So the question remains, "what is doing that every 5 minutes"?

[Mies]

Unfortunately the problem still exists.

 

Maybe searching on the lan is not the correct word, the spinner in running.

I don't know what in the background is happening, but when I disconnect the share, it does not happen anymore.

 

I think, that I have good quality devices like, Vu+, Synology, Netgear switches, etc.

Everything is running as it should be, no problems with SMB share from windows to windows or other devices.

 

I'm working in the IT and know more then a standard user, but this problem is not easy.

[WanWizard]

Like I've said a few times, you have something running that scans or reads all your files. And I still think Autotimer is the most likely culprit.

 

Restart Enigma in debug mode, wait until the problem occurs again, restart Enigma again in normal mode, and examine the created debug log. Or zip and post it if you want us to have a look...

[Mies]

Maybe, I'm on to something.
This, because the previous posts, forced me to think differently.

 

I have three Google Chromecast with Kodi installed with the enigma plugin, all in the bedrooms.
So I connect through port 80 and 8001 for streaming which works good but I never close Kodi at night.
Also I assumed that chromecast will close itself after a while because of the screen pictures.
I think that is not the case.

So I closed all Kodi's on every Chromecast.
I have to be careful in my predictions but it looks promising: Two days without spinner (unresponsive)

 

 

[Mies]

Until now i didn't have any spinners anymore.
let us wait a few days more but it looks promising.

[Jork]

If that is the case i still wonder why enigma is producing spinner locally? It still would be better to check the log just to have more info about this. I still think it can be avoided for this setup.

Edited by Jork, 6 September 2024 - 09:43.

[WanWizard]

Kodi enumerating all recordings through the web API, every 5 minutes ?