Problem, nur ein Volume funktioniert, Ruhe zeigt keine Bewegung!

[Adminelite 4]

My Server  AMD Ryzen™ V1500B quad-core 2.2 GHz.  32 GB RAM

hard disk: Seagate Exos X18 18 TB - intern - SATA 6Gb/s - 7200 rpm - Puffer: 256 MB

On Volume 1 there is only Emby System on it, there is no other program where I use it.

Volume 2 to 4 have my media data on it

 

When my friends family watch that only volume 1 goes high rest not why?

 

Edited January 6, 2022 by Adminelite
Fahler bei schreiben

[Luke 39685]

Hi there, can we please look at an example?

Thanks !

[Carlo 4549]

On 1/6/2022 at 3:04 AM, Adminelite said:

My Server  AMD Ryzen™ V1500B quad-core 2.2 GHz.  32 GB RAM

hard disk: Seagate Exos X18 18 TB - intern - SATA 6Gb/s - 7200 rpm - Puffer: 256 MB

On Volume 1 there is only Emby System on it, there is no other program where I use it.

Volume 2 to 4 have my media data on it

 

When my friends family watch that only volume 1 goes high rest not why?

 

I'm making an assumption so tell me if I'm wrong.  I don't think you created any raid volumes but instead use the drives individually?  If I'm correct then this is what I think from looking at that graph.

Unless you've changed the default location of the transcode folder I'd be almost certain that is what you see above for Drive 1 if your fiend is playing back Anime which isn't supported directly on their playback device.  If so, your server has to transcode it and Emby writes this out to the location you have set in the config or defaults to a transcode folder under it's own install directory. Besides the transcode Emby writes it's log on that drive as well as every other app including all the systems services.  It's going to be an active drive used most often if it's volume1.

Normally this isn't a problem because the assumption is all drives are pooled together using some type of RAID and share the load and combine the performance since you can read/write to multiple drives faster than a single drive.

Feel free to try setting the temp transcode location to a different location.  This is the option:

 

That's my best guess. I could be totally wrong and these are part of a raid array which would indicate an issue with that one drive (ie 5400 rpm and 3 other 7200 rpm).
Looks to me like the first.

 

[Adminelite 4]

16 hours ago, cayars said:

I'm making an assumption so tell me if I'm wrong.  I don't think you created any raid volumes but instead use the drives individually?  If I'm correct then this is what I think from looking at that graph.

Unless you've changed the default location of the transcode folder I'd be almost certain that is what you see above for Drive 1 if your fiend is playing back Anime which isn't supported directly on their playback device.  If so, your server has to transcode it and Emby writes this out to the location you have set in the config or defaults to a transcode folder under it's own install directory. Besides the transcode Emby writes it's log on that drive as well as every other app including all the systems services.  It's going to be an active drive used most often if it's volume1.

Normally this isn't a problem because the assumption is all drives are pooled together using some type of RAID and share the load and combine the performance since you can read/write to multiple drives faster than a single drive.

Feel free to try setting the temp transcode location to a different location.  This is the option:

 

That's my best guess. I could be totally wrong and these are part of a raid array which would indicate an issue with that one drive (ie 5400 rpm and 3 other 7200 rpm).
Looks to me like the first.

 

 

Hello my media are all x265 and without transcoding

 

Transcoding turned off stream original

have now tried the same effect with AFP.

with AFP my data is not showing on media

 

[Carlo 4549]

OK so you have transcoding turned off.

I meant to post this for you previously.  This is a week timeframe on my 920+.
We both are in the same range but have the opposite problem.  You have one drive high while I have way drive with far less utilization then the others.  Mine is explainable.
I had 4 WD Red 14TB NAS drives and wanted to start replacing them to go bigger so I replaced the first disk with an HGST 18TB Data Center Drive.  While the Reds are 65%-73% utilized the HGST seems to be around 14% utilized.  What makes this interesting is that my setup is an array and all drives should be working together doing "even" work.

I think it's obviously the HGHST is a much higher performing drive but another factor could be at play here on both our systems.  My Drive 1 is the newest drive added.  Right now it has 899 hours while the other 3 are around 3500 hours.  This drive likely far less fragmented.  Your first drive could be the most fragmented drive and they fore showing up with the most activity.

Have you used Storage Manager Volume tools to defragment your volumes?

Instead of trying to guess what's going on, let's find out. Go to control panel, file services.  For each protocol you have enable in each tab turn on logging for everything (all options). This will allow you to use the File Change log to see what's changed. Log Viewer will show you access to files. Resource Monitor Task menu services tab can be used to see which apps are writing and reading and can be sorted.

Last but not least take a look at the Emby Server log file (upload it here) to see what it's doing when you think there is a lot of activity.

 

[Adminelite 4]

On 1/8/2022 at 10:17 PM, cayars said:

OK so you have transcoding turned off.

I meant to post this for you previously.  This is a week timeframe on my 920+.
We both are in the same range but have the opposite problem.  You have one drive high while I have way drive with far less utilization then the others.  Mine is explainable.
I had 4 WD Red 14TB NAS drives and wanted to start replacing them to go bigger so I replaced the first disk with an HGST 18TB Data Center Drive.  While the Reds are 65%-73% utilized the HGST seems to be around 14% utilized.  What makes this interesting is that my setup is an array and all drives should be working together doing "even" work.

I think it's obviously the HGHST is a much higher performing drive but another factor could be at play here on both our systems.  My Drive 1 is the newest drive added.  Right now it has 899 hours while the other 3 are around 3500 hours.  This drive likely far less fragmented.  Your first drive could be the most fragmented drive and they fore showing up with the most activity.

Have you used Storage Manager Volume tools to defragment your volumes?

Instead of trying to guess what's going on, let's find out. Go to control panel, file services.  For each protocol you have enable in each tab turn on logging for everything (all options). This will allow you to use the File Change log to see what's changed. Log Viewer will show you access to files. Resource Monitor Task menu services tab can be used to see which apps are writing and reading and can be sorted.

Last but not least take a look at the Emby Server log file (upload it here) to see what it's doing when you think there is a lot of activity.

 

 

Hi

Now emby is even more stable with version 4.7.0.20 beta.  

 

 

 

 

old version 4.6.7.0 unstable synology

 

[Carlo 4549]

You really should take a look at Drive 1 to see why it's so much worse than the others.  Have you benchmarked it compared to the others, ran smart disk test on the volume?

[Adminelite 4]

On 22.1.2022 at 19:24, cayars said:

Sie sollten wirklich einen Blick auf Drive 1 werfen, um zu sehen, warum es so viel schlimmer ist als die anderen. Haben Sie es im Vergleich zu den anderen verglichen, einen Smart-Disk-Test auf dem Volume durchgeführt?

Smart disk test performed on the volume? no

Only when emby is running it gets that high

[Carlo 4549]

Try looking at the health of each drive.  I'm showing a quick test but you can run an extended test as well.

This is done in storage manage.

T

[Adminelite 4]

10 hours ago, cayars said:

Versuchen Sie, den Zustand jedes Laufwerks zu betrachten. Ich zeige einen Schnelltest, aber Sie können auch einen erweiterten Test durchführen.

Dies geschieht im Speichermanagement.

t

Good morning

did it. 

[Carlo 4549]

Looks like it's time to replace that drive as it's having a lot of seek errors.

[Adminelite 4]

18 hours ago, cayars said:

Looks like it's time to replace that drive as it's having a lot of seek errors.

it's funny that as soon as 35 people are looking it goes up

Jetzt:

 

hard disk:

 

disk 2 to 7 almost no movement hmm.  hard disk 1 only has programs on it, no videos

Edited October 23, 2022 by Adminelite
hard disk 1 only has programs on it, no videos

[Houfino 31]

Dne 1. června 2022 v 9:04 Adminelite řekl:

Hi..Can I ask? What is the program? Thank you

 

[Adminelite 4]

7 minutes ago, Houfino said:

 

Hello this is Synology Server

[Adminelite 4]

[Carlo 4549]

Looking at dev_0 & dev_6 you only have 1.2T and 54G stored on them.  What I would do is copy the data off these two to the other drives.

Without any other data on them remove them as a volume if this doesn't cause you a problem.  One may be your main drive so that's why I said copy vs move.

With both removed you should be able to connect them to a PC and run some windows utils on them to check them well and compare speeds between them.  You might be fine or if under warranty I'd get the one showing errors replaced.

Id then free up a 3rd drive by copying it over to the others as well.  Now put your 3 "fresh" drive in and set them up as Synology Hybrid RAID (SHR) with data protection for 1 drive fault tolerance.  That will give you some protection as well as speed things up a bit.  Use BTRFS file system with snapshot protection.

Now fully move the data from one of the other drives over to your new hybrid RAID.  When it's empty, add that drive bay to the hybrid RAID array.  Keep doing that until all drives are part of the hybrid array.

If your NAS supports NVME drives then I would add one or two of them as well so NVME can be used strictly for Emby which can really speed things up!  This requires a special setup but we have this written up as a PDF.

Carlo