Drive prevent from spin down


Recommended Posts

Hi @all,

I am new to Unraid and really loves it!

I have an Tower with 3 x 2 TB data drives and 1 x 2 TB parity drive.

My main goals with this setup: save all my data, plex and NextCloud.

 

After everything is setup and running really good, I mentioned that the parity drive and the first data drive will ever be online. 

Before installing NextCloud and plex I mentioned that ALL drives will spin down after some time. I guess that the docker's are preventing the array to spin down. 

 

Cause I don't find the exact problem in the documentation, I want to ask you to help. 

How could I spin down the array to save livetime of the disks (and some power) with active dockers?

 

I tried to understand the cache drive topic and guess, that this could help here. Maybe somebody could explan it a bit for me? 

I found this post, but don't get it all all https://forums.unraid.net/topic/46802-faq-for-unraid-v6/?page=2&tab=comments#comment-537383

Is cache only my option? Or do I need to manually copy files to cache drive?  

Link to comment
7 minutes ago, JulianB said:

Or do I need to manually copy files to cache drive?

A little confused by this question as the rest of your post makes me think you don't currently have a cache drive. If you have or get a cache drive it shouldn't be necessary to manually copy anything there, we can get mover to move them to cache.

 

Do you currently have a cache disk?

Link to comment

Hi Constructor, 

thanks for the fast help and answer. I have missed to say, that I am currently having no cache drive.

If a cache drive could help in this topic, I will buy one. 

 

So you say that the docker will keep the drives online. And when I will buy a ssd and put it into my tower, the docker could work on the ssd and the data array will completely spin down? If so, which config do I need to set in the config of the cache drive?  

Link to comment

I don‘t have that much Data that will be transferred to the Tower. Meaning that there May be some Images in the week and therefore i will Not Need the Cache. 

There is a bottlenec in the Drive capacity: i only habe 4 sata Ports on the Board and 1 PCI Slot. Yes, PCI and Not PCI-E. But it should work for one Cache Drive.

so let‘s say that i will only Need a Drive Cache for Reading.

is my Problem solved and the Hard drives will sleep? (Data Drive and parity Drive)

if so, which config Could I use? 

 

Thanks again for Helping 

—— 

edit: After a Short Look on Amazon and searching for a ssd, i have noted the prices of ssds... 240 gb for 40 €. Will this be enough for e.g. 50 gb of Write Cache and 190 gb of read Cache? 

Edited by JulianB
Link to comment
2 hours ago, JulianB said:

i will only Need a Drive Cache for Reading.

If you let your dockers use cache, as recommended, then it will be both reading and writing. The whole point is to get those writes off the array. You would just not be using them for user share writes if I understand your intention, which is fine.

 

2 hours ago, JulianB said:

Will this be enough for e.g. 50 gb of Write Cache and 190 gb of read Cache? 

This question suggests you have some fundamental misunderstanding of how cache is used.

 

That capacity will be fine though. Later you can add another to get some redundancy.

 

2 hours ago, JulianB said:

which config Could I use?

I don't really understand this question. As mentioned, there will be multiple steps involved in getting this setup.

Link to comment

Hi Constructor,

 

attached you will find the diagnostic file. What I did so far:

1. stopped the dockers (plex and NextCloud2)

2. stopped the array

3. added the new ssd as cache rive

4. created a btrfs partition on the cache drive

5. changed the share appdata to Cache only mode

6. started the array

7. moved (mv) the folder /mnt/disk1/appdata to /mnt/cache/.

8. now I made a mistake I think: I tried to move the folder under /mnt/user/appdata to /mnt/cache/appdata. After over 40 minutes I canceled the mv command. The folder was over 40 gb big in mnt/cache after that. However, I continued...

9. changed the docker App Config paths:

9.1  plex: from /mnt/user/appdata/binhex-plexpass/  to /mnt/cache/appdata/binhex-plexpass/ and

9.2 nextcloud2: from /mnt/user/appdata/nextcloud2/ to /mnt/cache/appdata/nextcloud2/ 

10. started both dockers

 

Current situation: Dockers are up and running. I could play all my music or watch my images in nextcloud. The ssd pushes it really forward! The speed increase was very good!

I am not quite sure If my drives now spin down. mmmh.. Everytime when I am looking in the webgui, the drives are up.

 

Thanks again for the help. And I really hope that I have not done too many mistakes ;-)

 

Edited by JulianB
Link to comment

I copied the last files... better the Mover copied it for me. 

After starting the Dockers again, I noted in the settings of the vm and docker, that the default location is on the normal user share and not in the appdata...

Do I also need to change it there? 

 

After I have started the dockers again, they are running all good. But the main drives still stay online (parity and disk1). 

You have any idea what to do? If the dockers might be a problem, i could delete them and re-create... some work but it could be done (It takes much time to import all photos and creating thumbs of them).

 

Link to comment
1 hour ago, JulianB said:

I copied the last files... better the Mover copied it for me. 

After starting the Dockers again, I noted in the settings of the vm and docker, that the default location is on the normal user share and not in the appdata...

Do I also need to change it there? 

 

After I have started the dockers again, they are running all good. But the main drives still stay online (parity and disk1). 

You have any idea what to do? If the dockers might be a problem, i could delete them and re-create... some work but it could be done (It takes much time to import all photos and creating thumbs of them).

 

Not quite sure what you mean?   The location /mnt/user/appdata and /mnt/cache/appdata are two ways of potentially referring to the same location.   The /mnt/cache/appdata is looking explicitly on the cache disk for the appdata folder, while the /mnt/user/appdata is looking on all array disks and the cache disk for such a folder.    If you have the ‘appdata’ share set to User Cache=Prefer they are effectively the same thing as with that setting mover will (if necessary) move any files belonging to that share from the array to the cache.

 

The advantage of using the /mnt/user/appdata path is that if you do not have a cache disk it is still valid (as it can then use an array disk), while if you try and use /mnt/cache/appdata without a cache disk present you end up writing to RAM which can cause all sorts of problems.

Link to comment

OK, attached you will find my new diagnostic. 

When you will be able to find a wrong config, that would be great. If not I will try to throw away the dockers and set it up again. 

 

@itimpithe only problem i have is that the main drives will not spin down. It is good to know that there are many problems when tower will use ram. But hopefully it is only uising the ssd. 

 

Edited by JulianB
Link to comment

Looks like the shares that belong on cache are all on cache now so that looks good. You only have a couple of other shares, and they have files on disk1, isos and another with name starting with D. If you have VMs or dockers using those shares then that might explain the drives not spinning down.

 

There are a few plugins that might help figure it out. Go to the Apps page and take a look at Dynamix Active Streams, File Activity, and Open Files plugins.

  • Like 1
Link to comment

I installed some of them and found the bad program... it is/was nextcloud that prevented the spin down of my drives. 

I couldn't say enough thank you for the help guys! This helped me a lot. There is really much to learn about Unraid. But I really like it a lot. 

 

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.