Pliind Posted November 30, 2019 Share Posted November 30, 2019 (edited) Hi, Quite new to Unraid and during my month with it I can't tell you how many times I've reinstalled Sonar / Radarr / Plex Media server because of database corruption. Setup: Unraid: 6.7.2 Cache: not using cache drive anymore since Sabnzbd froze the whole drive while unpacking etc, so all appdata inc docker.img is located on a seperate SSD. Appdata share has cache turned off. Issue: Sonar and Radarr get "Task Error: database disk image is malformed database disk image is malformed" and Plex Server starts acting up by ignoring one of my libraries and just being really buggy. I've heard that sharing your appdata over smb is not ok, so after my last "meltdown" i turned off smb and did a clean install of above apps, i though i was in the clear but yet Sonarr has gona corrupt on me. I've tried basicly all different containers aswell, binhex-sonarr, plex from linxuserver, plex binhex-plex, same result after a couple of days. (restoring config on sonarr and radarr is ofcourse tried, but its just a short term fix as the db will get corrupted again) Any ideas to what i can do so i can go on with my life? Edited November 30, 2019 by Pliind Quote Link to comment
Squid Posted November 30, 2019 Share Posted November 30, 2019 Update to 6.8.0-rc7 Quote Link to comment
Pliind Posted November 30, 2019 Author Share Posted November 30, 2019 Thanks, updating now. hope this works Quote Link to comment
Pliind Posted November 30, 2019 Author Share Posted November 30, 2019 I'm up and running now, but Plex will not start (Bad parameter) (Error response from daemon: Unknown runtime specified nvidia." . as I found another member on the forums has the same issue, removing "--runtime=nvidia" will make Plex Run, but GPU-transcoding will not work.. Quote Link to comment
Squid Posted November 30, 2019 Share Posted November 30, 2019 You were running the custom version of unRaid with the nvidia plugin. You need to go to the plugin and update from within there now. Not sure which rc version it supports. Quote Link to comment
trurl Posted November 30, 2019 Share Posted November 30, 2019 Are you mapping the Unassigned Device r/w slave? Quote Link to comment
Pliind Posted November 30, 2019 Author Share Posted November 30, 2019 yeah i saw that now, up to 6.8.0rc5 *gulp* Quote Link to comment
Pliind Posted November 30, 2019 Author Share Posted November 30, 2019 3 minutes ago, trurl said: Are you mapping the Unassigned Device r/w slave? the separate SSD is added as an part of the array Quote Link to comment
trurl Posted November 30, 2019 Share Posted November 30, 2019 12 hours ago, Pliind said: the separate SSD is added as an part of the array SSDs are NOT recommended in the parity array. They can't be trimmed, and there is some question whether some implementations might invalidate parity. Quote Link to comment
Pliind Posted November 30, 2019 Author Share Posted November 30, 2019 8 hours ago, trurl said: SSDs are NOT recommended in the parity array. They can't be trimmed, and there is some question whether certain implementation might invalidate parity. Hi! Could my corrupted databases have anything with my config to do? excuse my ignorence on this, it's been 2 weeks of trial and error.. How would i migrate my ssd out of the array without having to begin from scratch? so tired of reconfiguring everything (deconz and plex dont seem to like being restored with CA backup restore app. ) Quote Link to comment
trurl Posted November 30, 2019 Share Posted November 30, 2019 Go to Tools - Diagnostics and attach the complete diagnostics zip file to your NEXT post. Quote Link to comment
Pliind Posted November 30, 2019 Author Share Posted November 30, 2019 beast-diagnostics-20191130-2025.zipAttached Quote Link to comment
trurl Posted November 30, 2019 Share Posted November 30, 2019 Do you have any VMs? Your appdata, domains, and system shares have files on the array and they are not set to cache-prefer or cache-only. Ideally you want all of these shares to be completely on cache and stay on cache, so your dockers and VMs will perform better and so they won't keep your array disks spinning. Your SSD disk4 could be put in a pool with the nvme disk that is already cache, but unfortunately both disks would have to be reformatted as btrfs to be in a pool. 1 hour ago, Pliind said: it's been 2 weeks of trial and error So you have quite a few things that aren't configured ideally. Personally, I would start over except for the contents of your other user shares, which should wind up on the array, and maybe preserve your appdata so you could try to reuse it when you add your dockers again. It looks to me like you tried to run before you knew how to walk as the saying goes. Take things a little at a time and get each thing working well before trying to add more. Quote Link to comment
trurl Posted November 30, 2019 Share Posted November 30, 2019 Go to Shares - User Shares, click Compute All, wait for the results, then post a screenshot. Quote Link to comment
Pliind Posted November 30, 2019 Author Share Posted November 30, 2019 49 minutes ago, trurl said: Do you have any VMs? Your appdata, domains, and system shares have files on the array and they are not set to cache-prefer or cache-only. Ideally you want all of these shares to be completely on cache and stay on cache, so your dockers and VMs will perform better and so they won't keep your array disks spinning. Your SSD disk4 could be put in a pool with the nvme disk that is already cache, but unfortunately both disks would have to be reformatted as btrfs to be in a pool. So you have quite a few things that aren't configured ideally. Personally, I would start over except for the contents of your other user shares, which should wind up on the array, and maybe preserve your appdata so you could try to reuse it when you add your dockers again. It looks to me like you tried to run before you knew how to walk as the saying goes. Take things a little at a time and get each thing working well before trying to add more. Hi, No VMs at the moment. So i did start of with btrfs on the cache pool but as soon as sabnzbd was unpacking all other dockers froze, so thats why i moved one out of cache. As i have a dedicated ssd for my dockers, is my best bet moving my ssd from the array to an unassigned drive? The cache got enough to do temporarily holding media before getting moved to the array. Quote Link to comment
trurl Posted November 30, 2019 Share Posted November 30, 2019 1 minute ago, Pliind said: As i have a dedicated ssd for my dockers, is my best bet moving my ssd from the array to an unassigned drive? Some people do it that way but you have to take a little more care when setting up your dockers. Do you have a good understanding of docker volume mapping? Your screenshot is about what I expected, but I just wanted to make sure because your flash drive has a lot of files in config/shares (where the user shares configurations are stored) that don't appear to be used and their (anonymized) names seem very peculiar. Possibly you accidentally created a lot of user shares at some point by copying a lot of things to the top level of a disk. Quote Link to comment
Pliind Posted November 30, 2019 Author Share Posted November 30, 2019 7 minutes ago, trurl said: Some people do it that way but you have to take a little more care when setting up your dockers. Do you have a good understanding of docker volume mapping? Your screenshot is about what I expected, but I just wanted to make sure because your flash drive has a lot of files in config/shares (where the user shares configurations are stored) that don't appear to be used and their (anonymized) names seem very peculiar. Possibly you accidentally created a lot of user shares at some point by copying a lot of things to the top level of a disk. I think i do? Ive set all my dockers to install in disk4 rather than user share. And I understand how to map paths to inside of docker containers. when I transferred files from my old server to this one i made a mess of everything because of not really knowing what was happening and basicly filling my cache and making my dockers seaze. When i was working out the relation between sab sonarr and radarr and getting my files moved i might have placed downloads in my top level at first now that i think of it. how would i go about cleaning my flash drive and what would you suggest i do / read to understand why im making my server sad 😛 Quote Link to comment
trurl Posted November 30, 2019 Share Posted November 30, 2019 On your flash drive, in config/shares, there should be a .cfg file named for each user share, such as appdata.cfg. If you want to get rid of those files that aren't used anymore, you can just delete them and only keep the .cfg files that correspond to the user shares you currently have as shown in that screenshot. It would make your diagnostics somewhat less trouble to wade through. I've never run my dockers from an Unassigned Device so I'm not entirely sure how that works. I assume you would want the UD to automount but I don't know if there is something else needed to make sure the docker service doesn't start until after that mount. Maybe someone else can chime in here. 1 Quote Link to comment
Pliind Posted November 30, 2019 Author Share Posted November 30, 2019 Thanks, what app should i use for that, krusader wont let me in, and sharing it on smb might not be appreciated by the system? Quote Link to comment
trurl Posted November 30, 2019 Share Posted November 30, 2019 Simplest is to just shutdown and put your flash drive in your PC. 1 Quote Link to comment
Pliind Posted November 30, 2019 Author Share Posted November 30, 2019 Gotcha, will do! Quote Link to comment
Recommended Posts
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.