trurl

👍

Your user share named "all" still has files on cache. Do you know how to see what those are?

You might consider replacing parity with a larger drive and reusing parity as a data disk. That way you will be able to use larger drives in your array, whether as replacements or in new data slots. 8TB seems the be the sweet spot at the moment as far as TB per dollar.

It is impossible to move data from fast SSD cache to the slower HDD parity array as fast as you can write to fast SSD cache, no matter how large your cache is. You just need to reconsider how you are using cache. First of all, don't cache the initial data load. Cache only gets in the way because Some people even do the initial data load without parity since parity slows down writes. After the initial data load, consider whether you really need fast writing to some of your shares. Many of the writes to my server are from unattended processes. Since I am not sitting there waiting for them to complete, no need to cache them. Finally, mover is intended to run during idle time because Instead of being concerned with how much you can write in an hour, it would really be better to consider how much you might write to cache (and how much you might instead write directly to the array) during the day so it can be moved to the array at night when you aren't writing new files to cache. If you want more specific advice on how to avoid filling cache, post your Diagnostics. Tools - Diagnostics, attach complete Diagnostics ZIP file to your NEXT post in this thread.

Disk firmware has no way to know about filesystems, that is at the OS level.

yes just include the next disk in the share. That is similar to what I do except I don't use fillup.

👍 yes. As for getting someone else to check your connections I will let you decide. Since you don't need the capacity of a third drive yet you might wait to add that later. I always say each additional disk is an additional point of failure.

How are you accessing it? I don't see WireGuard plugin in your diagnostics. If using OpenVPN docker, for example, then there is also going to be some trouble fixing those other things I mentioned since you will have to disable dockers.

I'm a windows guy also. This is about using Unraid, not the underlying linux OS. The split level setting is about a user share, not about a drive. In fact, the whole point of split level is to keep certain files together so they don't go to other drives. For example, suppose you have a music collection, and you are old-school and like to listen to "albums". You are playing an album, and the first several songs are all on the same disk. So it plays the songs from that disk, but since the other disks aren't being used, they are spun down. Now it gets ready to play the next song from the album, but that song has been allowed to be split to another disk. Since that other disk is spun down, you have to wait for it to spin up before the song can begin playing. Same principle when binge-watching TV series. Some people will prefer to keep all the files for a specific TV series on the same disk so there is no spinup delay between episodes. I don't care about split level, mostly because I have my whole music collection restricted to a single disk anyway, and I don't care about TV much. So for most user shares, I just set split level to allow splitting anything. You have split level set so it WON'T split anything. Instead, you have told it that you will manually create new folders on other disks when you want files to go to other disks. You can click on the label for any setting in the webUI to toggle help for that setting. For simplicity, try changing it to "Automatically split any directory as required".

Change the split level on your TV share!!!

yes, did you understand my reply? So does this mean you are good now?

Why are you ignoring us and just repeating yourself ??

Mover can't move open files, and the files are open all the time when Docker and VM services are enabled. Also, your domains share is actually set to be moved from cache to the array. There are some specific things that need to be done to get this all done. Let me know when parity rebuild is done and we can work on it.

It makes it impossible to coordinate replies

You have to either change the split level or manually create folders on disk6 for new TV share files.

@itimpi didn't realize you had been so busy with the wiki. Great job. I need to link it more often.

But also note that using an earlier version, then upgrading to 6.9.x, while it would still support the old alignment, will require reformatting to take advantage of the new alignment.

Let me know if you want to work on those other things.

Looks good. You might consider moving domains share to cache also but I didn't mention it since you didn't have VMs enabled. Depends on how you configure VMs to use storage.

Why have you allocated 50G to docker.img? 20G should be more than enough unless you have something misconfigured. Also your appdata, domains, system shares have files on the array. This will make dockers/VM keep array disks spinning since these files will be open, and docker/VM performance will be impacted by parity updates. But, that isn't what you were asking about. SMART for all disks looks fine. My guess is connection issues. These are common when mucking about in the case. Check all connections, SATA and power, both ends, including splitters. Maybe consider changing SATA cable.

I usually follow up with a noncorrecting parity check to verify that there isn't some other problem. The only acceptable result is exactly zero.

Unraid doesn't have wifi so it requires an ethernet connection to the network. What do you mean about connecting a monitor by ethernet to the server? By "monitor" do you actually mean a display monitor, or do you mean another PC? You can connect a display monitor directly to your server using whatever video capabilities and cable your server hardware and display monitor support. Ethernet isn't normally used to connect a display monitor to a computer. Maybe something I am not understanding about your question.

I assume that is either writing to cache, or no parity installed. Even so, I thought the scenario you were considering now was going from old server to pc to new server. You could go directly between servers and leave out the PC. Or just do one disk at a time directly on the new server with Unassigned Devices and leave the network out completely. You wouldn't even need the old server running and could just go with your already licensed flash drive on the new server.

Here is the 6.9beta25 thread: https://forums.unraid.net/bug-reports/prereleases/unraid-os-version-690-beta25-available-r990/ I actually did the multiple pool setup on beta22: https://forums.unraid.net/bug-reports/prereleases/unraid-os-version-690-beta22-available-r955/ But when the new partitioning for SSDs came out I redid them on beta25. So, I have moved appdata, domains, system shares around quite a bit lately. The multiple pools especially came in handy for the repartitioning, since I just moved them manually with mc between the SSD pools so I could repartition. Much faster than getting Mover and the parity array involved.

@Hoopster was the inspiration for my latest (re)build: https://forums.unraid.net/topic/93149-mitx-motherboard-which-can-support-7-drives-on-board/