How to re-use unused cache disk

[SliMat]

Hi all

 

I had a 1.2Tb cache disk which I have never used. Now my array is full and I need space I stopped the array and unassigned it as cache disk. Restarted the array and it warned me it would remove the cache disk assignment... which I did.

 

I then expected the disk to be available to add to my array - I added an extra slot (number 19)... but the 1.2Tb disk is listed under "Unassigned Devices" and I can work out how to assign it to the array as its not listed as available to allocate to the slot.

 

I know I am missing something simple - but can anyone offer any pointer?

 

Thanks

[trurl]

Do you mean you aren't using a cache disk at all, or just not that cache disk? Might be better to reconsider if you don't have a fast pool unless you are only using NAS functionality.

[SliMat]

Thanks @trurl

 

I fitted a 1.2Tb cache disk, but have never assigned anything to it.

 

At the moment the server is only being used for Plex, Nextcloud and a couple of webserver VMs for some WordPress websites... I retired my Exchange server and went with Office 365. However at the moment I have 14Tb of media in an 18 disk array (with 2 x 2Tb parity disks). Recently the server is down to a few Gb of space and last night I couldnt watch anything on Plex as it said that there wasnt enough disk space to convert the media. I deleted a couple of box sets and restarted it as a quick fix.

 

So, I got to thinking that as I have a spare 1.2Tb disk assigned as a cache and not actually doing anything I would add it to the array to give me another 1.2Tb of media space. My thoughts are that if I dont knowingly have any speed issues then I obviously dont need a cache disk - especially as its been installed for a couple of years and not been used for anything!

 

Any thoughts, comments?

[JonathanM]

13 minutes ago, SliMat said:

 

At the moment the server is only being used for Plex, Nextcloud and a couple of webserver VMs for some WordPress websites..

Which disk are they using? VM's and containers typically perform very poorly when run from an array disk.

 

The default configuration of Unraid uses the cache for containers and VM's, you would have had to manually put them elsewhere.

[SliMat]

Hi @JonathanM

 

Yes, I configured them manually... eg;

 

 

To be honest I havent noticed any real issues with performance - but to be fair I've never moved them to a cache disk to compare.

 

I do have some spare 1.2Tb disks, so perhaps I should send them to the datacentre where my server is installed and get them to add them as I have about 5 or 6 spare disk bays - so could add them and then re-assign this 1.2Tb disk as a cache disk.

 

From memory the reason I never used the cache disk for VMs or dockers is because of the risk of losing all if the disk failed.

 

Thanks

[JonathanM]

20 minutes ago, SliMat said:

Yes, I configured them manually... eg;

 

That location could be located on the array or on the cache, /mnt/user encompasses both locations.

 

Shares, compute all will tell you which actual disk holds what. It will take a few minutes, just leave it on that page and refresh the page after a bit.

[SliMat]

I did look on the cache disk (before I unassigned it as cache disk) and the only folder on there was an old Plex one from when I ran Plex as a docker. It now runs in a VM. So, I deleted the redundant /appdata/plex folder and all still works... also, the server is active right now and everything is working with the cache disk removed.

[trurl]

1 hour ago, SliMat said:

18 disk array (with 2 x 2Tb parity disks)

So you have a large number of small disks. I always recommend fewer larger disks. Each additional disk is an additional point of failure.

 

That doesn't match your sig at all by the way. Unfortunately sig viewing is disabled by default on this forum software. You can enable sig viewing and edit your sig by clicking the dropdown from your username at upper right and going to Account Settings.

 

If that 1.2TB disk we are discussing is a spinner, not really the best thing for fast pool anyway.

 

 

[JonathanM]

3 hours ago, SliMat said:

the 1.2Tb disk is listed under "Unassigned Devices" and I can work out how to assign it to the array as its not listed as available to allocate to the slot.

Is it mounted in Unassigned devices?

[SliMat]

Hi @JonathanM

 

No, its not;

 

 

Which is why I find it odd that its not listed as available to put in a spare slot.

[SliMat]

1 hour ago, trurl said:

So you have a large number of small disks. I always recommend fewer larger disks. Each additional disk is an additional point of failure.

 

That doesn't match your sig at all by the way. Unfortunately sig viewing is disabled by default on this forum software. You can enable sig viewing and edit your sig by clicking the dropdown from your username at upper right and going to Account Settings.

 

If that 1.2TB disk we are discussing is a spinner, not really the best thing for fast pool anyway.

 

 

Hi @trurl

 

Yes, sorry that I havent updated my sig in ages and forgot The machine is a DL380 with 25 x 2.5" SAS disk slots. Currently I have 2 x 2Tb, 5 x 1.2TB, 2 x 1Tb, 2 x 600Gb & 6 x 500Gb SAS disks. Then I also have 2 x 2Tb disks for parity.

 

Yes, the 1.2Tb disks is a spinnig SAS disk, again another reason why I was planning to add to the array. I did explore an SSD cache disk and needed at least a 1Tb SSD as one VM alone was about 300Gb. This was my Exchnage server which is now redundant... so I may be able to get away with a smaller SSD as a cache to keep costs down. I havent looked into this recently... so might re-explore.

 

So, in the meantime its just finding out why the 1.2Tb ex-cache disk isnt showing as available to add to the array. I did wonder if its because its formated BTRFS and not XFS... but I would expect UnRAID to identify the disk and allow a re-format.

[trurl]

maybe diagnostics would tell something

[SliMat]

Thanks @trurl

 

Is it best to attach the whole log zip file... or is there a specific file within the zip file to attach?

[trurl]

Complete Diagnostics ZIP, attach to your NEXT post in this thread 

[SliMat]

Thanks @trurl

 

Attached.

dl380p-rack-diagnostics-20220324-0924.zip

[SliMat]

Hi @trurl

 

I was wondering if the diagnostic file gave you any clues why I cant add the former unused cache disk in to the main array?

 

Thanks

[itimpi]

The diagnostics show multiple attempts to log into your server from a variety of internet addresses.   Do you have your server exposed to the internet (or in your routers DMZ)?  Unraid is not hardened enough to be directly exposed to the internet.  If you want to be accessing Unraid from the internet then you should be using either the My Servers plugin or be securing the link using a VPN (such as the built-in WireGuard) or an equivalent.

 

Because of the numerous ongoing messages showing attempts to log into your server from the internet i find it is not practical to try and find any relating to your drives.

[SliMat]

Hi @itimpi

 

Yes, this UnRAID server is in a datacentre - but it is behind a firewall (Mikrotik router).

 

I do have Plex running which has serveral remote users that access it. I also have Resillo Sync running to synchronise files between two servers. I did have an Exchange Server running which was compromised as a result of a Microsoft vulnerability last November!

 

Also running NextCloud - but this is only used by me and my brother for backing up work files.

 

I also have a webserver/FTP server which allows suppliers to upload their price lists every hour too.

 

I thought having the whole UnRAID box only accesible via a VPN would stop users accessing NextCloud/Plex?

 

I'll have a read up on "My Servers" / "Wireguard" as I dont know anything about these.

 

Would the above explain the remote access attempts you mention?

Edited April 4, 2022 by SliMat

[itimpi]

37 minutes ago, SliMat said:

I thought having the whole UnRAID box only accesible via a VPN would stop users accessing NextCloud/Plex?

Do not think this is the case but I am not sure  

37 minutes ago, SliMat said:

Would the above explain the remote access attempts you mention

 

there are thousands of attempts so it seems unlikely but I could be wrong.  If you look at the syslog files in the 'logs' folder within the diagnostics you. will see what I mean.

 

[SliMat]

Hi @itimpi

 

Yes, I had started to look at these and I suspect it was from when my Exchange server was hosted on UnRAID as this became compromised and it looks like someone has been attempting some sort of brute force attack - but, unless I am mistaken I dont see anything recent. I switched to Office365 in January to release me the burden of trying to maintain a creaking Exchange server 🙂

 

Any pointers on securing UnRAID greatly received!

[itimpi]

29 minutes ago, SliMat said:

Hi @itimpi

 

Yes, I had started to look at these and I suspect it was from when my Exchange server was hosted on UnRAID as this became compromised and it looks like someone has been attempting some sort of brute force attack - but, unless I am mistaken I dont see anything recent. I switched to Office365 in January to release me the burden of trying to maintain a creaking Exchange server 🙂

 

Any pointers on securing UnRAID greatly received!

I must admit I had not looked at the dates  

 

when I look now the latest log entry seems to be Feb 18th which is a bit suspicious.    I would have expected to see some recent entries from your attempts to get the new drive to be recognised?   Maybe it is worth rebooting the server and then getting new diagnostics that are up-to-date?

[SliMat]

Quick Update - during the reboot process I thought I'd play with the newly free'd up cache disk - so I re-assigned it as a cache disk then un-assigned it (again) and now it appears as assignable in the array 🙂

 

So... I'll keep checking for pointers on security practices. I currently have 6.8.3 installed as I was trying to use the SAS Spindown plugin which didnt work on my machine with 6.8.3... so I am going to upgrade to 6.9.2 now and play with "My Servers" too.

Edited April 5, 2022 by SliMat
typo