trurl

Why run plex in a VM? There are several plex dockers and they work well.

done

I would say that your main problem is you have a very large number of disks in the parity array, and these are almost completely full. Personally if I were going to have an array with that many disks, I would be running dual parity. You also have way more user shares than seems necessary. That large number of user shares, combined with your nonstandard way of using them for you docker configuration, makes it more trouble to try to make sense of them though, especially since the share names are anonymized in the diagnostics. Just looking at a few of those, it doesn't seem like you have bothered to set one of the most important settings for the shares, Minimum Free. And since your disks are so full it is likely that you will try to overfill your disks because there is no Minimum for your user shares. Have you ever had any errors due to lack of space? I don't want to look at all those SMART reports and that isn't your purpose for this thread anyway, but I still want to know. Are any of your disks giving SMART warnings on the Dashboard? What is your plans for increasing capacity? I know you are really just trying to get your dockers working again. but your whole server setup is teetering on the edge at this point and those things need to be taken care of before trying to get your dockers setup correctly and working again. Do you have any VMs?

And I missed this one No it is not like RAID5, what he was describing is one of the things that makes Unraid very different from any RAID.

Don't really understand that first sentence. And it was in response to a quote that had some things wrong as far as how parity works. As for those more experienced people, some of them have already replied on this thread. Parity can detect and correct a single bit error. Dual parity can detect and correct 2 bit errors. Parity is very common in computing and communications, and it is always conceptually the same though the specific implementation may differ in some respects. If you have a group of bits, determine if their sum is even or odd then include an additional bit that reflects that even or odd. Then if a bit is missing you can figure out what it was based on whether it should be even or odd. I already posted a link to the wiki on parity above.

Doesn't have to be the largest, each parity drive has to be at least as large as the largest single data drive.

Read his response on this same page

Flash drive has disconnected. Don't know if there is anything actually wrong with it though. Are you using a USB2 port to boot? You should.

And if you think about it for a moment, it would not be possible for parity by itself to recover a file from any possible disk that might fail. For that to happen, parity would effectively be a backup of the complete array. There is no magic that would allow it to have the capacity for that.

post that system log

Your share 1 scenario doesn't really make sense. Guests are anyone, so if you don't give user 2 access, they get guest access. Even if there were a way to specify that, user 2 could just login as JoeBlow and they would be a guest. So no point in making it possible to give a specific user less access than JoeBlow

Reviewing the thread, I wanted to go back to this: Best if you only work with your server in a single browser tab on a single machine. Has mover finished yet? If so, post a new diagnostic.

Not clear what specifically you had wrong with your configuration, but many people don't have these problems. Perhaps if we could see how you had things setup and what was happening at the time we could have figured things out. Also not clear how formatting cache is going to solve anything. As long as cache is not corrupt things can be moved, copied, or deleted from it. We can rework your user shares and docker setup then work through setting up individual dockers. You've been on the forum for several years. Do you have a pretty good understanding of how to setup user shares and dockers yourself, or were you just relying on others to tell you what to do?

There isn't really the concept of "groups", if by that you mean user groups. There is the ability to control access to network shares for each user separately. Have you read the Product pages, linked at the top of every forum page? There is also this wiki Overview: https://wiki.unraid.net/UnRAID_6/Overview

I'm running 24/7 with non ecc. Don't understand this question. The whole point of a NAS is to save data. What sort of info are you thinking might be lost if you shutdown?

I assume you are referring to my sig. unSERVER is the name of my Unraid server. The "default" name for any Unraid server is "tower", but you can change that in Settings - Identification. The rest of that line in my sig is just my specs in very abbreviated form, each section separated by : is enough for a google search to identify the hardware component. And I also have useful links in my sig. Many other users have useful links in their sigs as well, but... For anyone else that might be reading, unfortunately, viewing sigs is disabled by default in this forum software. To enable viewing sigs and edit your own sig, click the dropdown beside your user name at the top right of a forum page, and select Account Settings. And if you had a sig in the old forum software, that sig was imported and you still have it even though you might not be aware of it. It is probably outdated so you should take a look at your own sig.

I'll answer in a slightly different order since it makes more sense that way. Yes each disk is an independent filesystem containing complete files that can be accessed on any linux. Each disk is an independent filesystem, so each disk contains complete files. But folders can span disks, and these spanned folders are presented as network shares. 1 (or 2) parity disk allows Unraid to calculate the data for 1 (or 2) missing disk to be calculated from parity plus all remaining disks. Here is the wiki on single parity: https://wiki.unraid.net/UnRAID_6/Overview#Parity-Protected_Array Dual parity is similar but the parity2 calculation is independent and somewhat more complicated. But, parity is NOT a substitute for backups. Plenty of ways to lose files besides disk failure, including simple user error. You must always have another copy of anything important and irreplaceable. Not familiar with that solution so I will just mention some advantages Unraid has over RAID. In addition to each disk being an independent filesystem which can be accessed by itself on any linux, Unraid also allows different sized disks to be mixed in the parity array, and you can easily replace or add disks without rebuilding the entire array. The tradeoff vs RAID is exactly that each disk is an independent filesytem, there is no striping so each file can only be accessed at single disk speed instead of striped RAID where multiple disks are accessed in parallel. Writes are actually somewhat slower than single disk speed since Unraid maintains parity in realtime. There is also cache or cache pool, optionally SSDs, which can allow faster I/O, including writes which are later moved to the slower parity array.

Closed and locked the report after copying to General Support. Very unlikely to be an actual bug, and the report doesn't follow the bug report guidelines anyway. https://forums.unraid.net/bug-reports/stable-releases/report-guidelines-r68/ If it does turn out to be a bug we can reopen. For now we can try to address it here. Did you read these upgrade notes, linked in the Release Announcement? https://forums.unraid.net/topic/66327-unraid-os-version-650-stable-release-update-notes/

Closed and locked this after copying to General Support. Very unlikely to be an actual bug, and the report doesn't follow the bug report guidelines anyway. If it does turn out to be a bug we can reopen

Changed Status to Closed Changed Priority to Other

Unraid does NOT run from flash drive. It unpacks the archives fresh from flash at each boot into RAM and it runs completely in RAM. Think of the flash archives as firmware.

It shouldn't trigger parity check, likely some issue with flash drive. See here: https://forums.unraid.net/topic/46802-faq-for-unraid-v6/?do=findComment&comment=801384

That doesn't show your cache anywhere near full. And appdata and system shares are on cache where they belong, and no other shares are currently using any of cache. So there is nothing for mover to do. What are you looking at the makes you think your cache is filling up?

Why do you want to reformat cache anyway? What is the ultimate problem you are trying to solve?