Help with HDD high read writes on 1 drive

[matto2494]

Hey guys, 

I am trying to figure out why one of my disks has really high reads and writes. my setup is as follows

• Partiy 6tb
• disk 1 6tb
• disk 2 3tb
• disk 3 3tb
• cache 240g SSD

 

To start i have stopped all dockers apart from nginxproxymanager and stopped my unifi VM.

Disk one is still climbing in writes and reads, granted it has slowed but it is still climbing.(The screen shot is of the reads and writes straight after clearing stats)

disk 2 and 3 spin down and very rarely come up.

 

i have this come up on the disk log info when i select disk 1 not sure what it is could anyone shed some light?

 

Hope this makes sense still learning about unraid even though ive been using it for 3+years

 

thanks in advance

[matto2494]

6hrs later and this is the amount of reads and writes i have on 1 drive.

 

[testdasi]

As always, Tools -> Diagnostics -> attach zip file.

[matto2494]

 

odin-diagnostics-20191209-2012.zip

[Squid]

You sure it's not Plex indexing the files or another app / vm running?

[testdasi]

Screenshots of docker and VM tabs please.

Preferably while you observe HDD activity.

[matto2494]

On 12/9/2019 at 10:40 PM, testdasi said:

Screenshots of docker and VM tabs please.

Preferably while you observe HDD activity.

I hope i have understood your request.

 

I rebuilt my cache drive last night and resetup all my dockers from scratch in a vain attempt to resolve the issue.

Still seems that the drive is constantly being written to.

 

 I think i have identified it to being something on the Appdata/domains/system shares.

I installed a cache drive a long time ago but obviously didnt do it correctly as there was still data on the array that should be on the cache.

Now i have fixed that the drive that is now being hammered is the cache instead of drive 1, so it seems to follow the appdata/domains/system folder

 

 

 

[itimpi]

4 minutes ago, matto2494 said:

Now i have fixed that the drive that is now being hammered is the cache instead of drive 1, so it seems to follow the appdata/domains/system folder

It is expected that those shares will get high traffic if running VMs or dockers which is one of the reasons having them on the cache is recommended (as well as overall performance of VMs and  docker containers).

 

The 'appdata' share will be where all the docker containers that are running are writing their variable data.

The 'domains' share will be where the vdisk for your unifi VM is located.

The 'system' share will be where the docker.img file is located that contains all docker containers.

 

As to whether the figures you are seeing are appropriate for your workload I have no idea.