Discs are not spinning down. Parity, Disk1, Cache was Plex was running

[chris_netsmart]

I hope someone can please explain this to me.

 

I am trying to get my disks to spin down, as and when they are no longer need, but it looks like Plex is doing something to keep the following disks up and spinning.

I have checked the folder structure and discovered the following.

 

Parity

Disk 1                    ISOS, Media, Multimedia, system, users

Cache                    Appdata, system

 

the other disks in my Raid have spun down.

 

I have also check that all the disks are set to default for the spin down, but I can’t work out what plex is doing.

 

Any ideals ?

Edited December 28, 2018 by chris_netsmart

[Can0n]

Cache and parity will never spin down as they are constantly in use....plex will do checks every now and then that might be causing disk 1 to not spin down best to check how often plex is set to check folders for new content

 

head to settings in plex then library check off the scan every 2 hours or whatever the timing is set to if its enabled

 

 

[chris_netsmart]

@CanOnFan

thanks for the information about the Cache and Parity,  

I have taken a screen shot of my settings

 

[Can0n]

yeah remove check next to scan my library periodically thats likely keeping your disk 1 from sleeping....keep in mind if the disk with the media you try to access is sleeping when you try to access it plex may spit out an error if it takes too long to spin up, also make sure if you stop playing on plex you completely exit the stream or it will hold onto the file again keeping the disk from sleeping

[kizer]

I also noticed you have a system on Disc 1. Are you sure that Plex is installed on your SSD not Disc 1?

[chris_netsmart]

Installed plex as default. I will go and recheck.

 

As i am still getting my head around this platform. 

 

I just looked at my plex advance and i see the following

 

/mnt/user/appdata/PlexMediaServer

 

How can i change this to the cache ? And not disk1

 

Edited December 28, 2018 by chris_netsmart

[Can0n]

26 minutes ago, chris_netsmart said:

Installed plex as default. I will go and recheck.

 

As i am still getting my head around this platform. 

 

I just looked at my plex advance and i see the following

 

/mnt/user/appdata/PlexMediaServer

 

How can i change this to the cache ? And not disk1

 

First shut down any dockers you are using, go to shares, and click on appdata share  under use Cache disk set it to only.

 

apply that go to the main page and click on Mover at the bottom it will take a while but once mover is done start up your plex docker and any other dockers you previously were running and should be fine

 

it will move everything from appdata to the cache (appdata is primarily dockers and should only bve run from your cache if you want to use decent speed.

 

if its currently spreading the Plex Database files across all your unraid spinning disks this explaisn why nothing can sleep

[chris_netsmart]

thanks for the advice, I have now moved the Appdata onto the Cache, and now I am getting 100% across all my CPU cores. and a click sound from my Cache drive.  

 

the Cache drive is a  WD Green 1TB HD, which is over 5 years old. "Testing only" I will leave the Plex running overnight and hopefully the CPU will reduce, and I will purchase a SSD tomorrow and replace my 1TB

[chris_netsmart]

All working fine. All my HD goes to sleep.

Also my CPU usage is down to 0% when there active streamimg.

 

Note. It takes about 10 mins from power on to having the system up and running with.a low CPU.

 

Thanks for your help

[Can0n]

On 12/28/2018 at 3:25 PM, chris_netsmart said:

thanks for the advice, I have now moved the Appdata onto the Cache, and now I am getting 100% across all my CPU cores. and a click sound from my Cache drive.  

 

the Cache drive is a  WD Green 1TB HD, which is over 5 years old. "Testing only" I will leave the Plex running overnight and hopefully the CPU will reduce, and I will purchase a SSD tomorrow and replace my 1TB

Cache drives should always be SSD’s otherwise there is no purpose for caching. 

 

 

[Can0n]

To check what’s eating your cpu

 

go to the built in terminal and type top and hit enter

 

paste the top 5 things to a reply

[Squid]

2 minutes ago, Can0nfan said:

otherwise there is no purpose for caching. 

Not true.

 

A spinner for a cache drive will be significantly faster than the array for writes, especially for docker containers, VMs  For accessing over the network, then even writes to the cache will be moderately faster, but not hugely.

[Can0n]

2 hours ago, Squid said:

Not true.

 

A spinner for a cache drive will be significantly faster than the array for writes, especially for docker containers, VMs  For accessing over the network, then even writes to the cache will be moderately faster, but not hugely.

News to me most videos i saw suggested SSD's for Caching for speed...if the spinner is still on same JBOD card or motherboard ports why is it faster as a cache than array? due to not writing to parity?

[itimpi]

3 minutes ago, Can0nfan said:

News to me most videos i saw suggested SSD's for Caching for speed...if the spinner is still on same JBOD card or motherboard ports why is it faster as a cache than array? due to not writing to parity?

Every write to the parity protected array actually 9nvolves 4 I/O operations (a read and write on both the parity and data disks) while a write to the cache is a single I/O.

[Can0n]

1 minute ago, itimpi said:

Every write to the parity protected array actually 9nvolves 4 I/O operations (a read and write on both the parity and data disks) while a write to the cache is a single I/O.

gotcha 

[trurl]

On 12/28/2018 at 2:40 PM, Can0nfan said:

First shut down any dockers you are using, go to shares, and click on appdata share  under use Cache disk set it to only.

 

apply that go to the main page and click on Mover at the bottom it will take a while but once mover is done start up your plex docker and any other dockers you previously were running and should be fine

 

it will move everything from appdata to the cache (appdata is primarily dockers and should only bve run from your cache if you want to use decent speed.

 

if its currently spreading the Plex Database files across all your unraid spinning disks this explaisn why nothing can sleep

Actually, this won't do what was intended. Mover won't touch cache-only shares.

 

Before giving you the correct procedure, lets figure out what your situation really is in more detail.

 

Go to Shares - User Shares and click Compute All. When it is done, post a screenshot.

[chris_netsmart]

19 hours ago, Can0nfan said:

To check what’s eating your cpu

 

go to the built in terminal and type top and hit enter

 

paste the top 5 things to a reply

as requested

 

 

[chris_netsmart]

16 hours ago, Can0nfan said:

News to me most videos i saw suggested SSD's for Caching for speed...if the spinner is still on same JBOD card or motherboard ports why is it faster as a cache than array? due to not writing to parity?

I am Looking at a M.2 chip. as I hear that this can my much faster then a SSD

[chris_netsmart]

3 hours ago, trurl said:

Actually, this won't do what was intended. Mover won't touch cache-only shares.

 

Before giving you the correct procedure, lets figure out what your situation really is in more detail.

 

Go to Shares - User Shares and click Compute All. When it is done, post a screenshot.

 

[trurl]

That's the right page but not the screenshot I wanted.

 

Did you click the Compute All button at the bottom of the User Shares section? After clicking Compute All it determines how much of each disk each user share is using and displays it. It will take some time to get the result.

 

[chris_netsmart]

29 minutes ago, trurl said:

That's the right page but not the screenshot I wanted.

 

Did you click the Compute All button at the bottom of the User Shares section? After clicking Compute All it determines how much of each disk each user share is using and displays it. It will take some time to get the result.

 

yes I did leave it for some time, I will kick it off and leave it over night. hopefully it will have the info you that

[trurl]

You may have to refresh the page after a while. I just did it on mine and it didn't really come back with the results until I refreshed.

[chris_netsmart]

ok, I have just returned to have a look at this and I see that my Disk1 is still active, after I 15 mins.

I have check the Appdata share and the Disk1 location, and I am happy to say that the Share is set to Cache only, and the Folder on disk1 is emtpy.

 

the only Docker that is running at the moment is Plex.

 

the folders that are on the Disk1, which are not mine are appdata , isos, and system.

I also have attached a copy of my computed shares

 

any ideals on why my disk1 is still spinning ?

Edited January 13, 2019 by chris_netsmart

[trurl]

If you aren't using VMs then isos share doesn't matter and it is empty anyway. You should delete the empty appdata folder on disk1.

 

system share exists only on disk1, it isn't empty and that is probably where you have it set to put your docker image, which is where the actual code for each of your containers resides.

[chris_netsmart]

28 minutes ago, trurl said:

system share exists only on disk1, it isn't empty and that is probably where you have it set to put your docker image, which is where the actual code for each of your containers resides.

So what you are saying is that it is normal for disk1 to be spinning ? And it will only sping down when, none of the dockers are updating.