Janddy

I was on 6.9.2 so I just upgraded to 6.10.3 and tried adding the URLs again. The icons updated and then survived a force update and a server restart so it looks like that fixed it. Thanks for the help, I can't believe it was that easy in the end! I should have upgraded sooner.

Hi, can anyone help me fix the icons on a couple of my docker containers? They are always reverting to a question mark no matter what I seem to do. After looking for solutions in the forums and google etc. I have tried: Changing the icon URL in the container template to a URL that contains the icon I want. This will work for a bit, and then revert back to a question mark after a couple of hours. Deleting the images from /var/local/emhttp/plugins/dynamix.docker.manager/images, adding the URL back to the container template, and force updating. Deleting the images from /var/lib/docker/unraid/images and copying the correct image into that folder, then deleting the old image from /var/local/emhttp/plugins/dynamix.docker.manager/images. After refreshing my browser the image appeared, but then after force updating the container it went back to a question mark. Deleting the images from /boot/config/plugins/dockerman/images. (There is no image in here for one of the containers). Restarting Unraid after trying the above steps. Clearing cache in browser and also viewing the docker page from a different browser. Using images at different URLs. If it matters, the problem containers are snipe/snipe-it and atribe/apcupsd-influxdb-exporter. The icons I am trying to use are snipe-it and apcupsd-influxdb-exporter. I have been having this problem for nearly a year and every time I look at my dockers I see the question marks begging to be fixed, but every time I sit down determined to fix it I come away with nothing. I'm at a loss - does anyone have any ideas on what to try?

Ha, didn't even think of that. I suppose that would work. May resort to that if needed. Interesting. I assume that would be something like this, or this? Are there other benefits of using an add on sata controller, apart from being easier to plug in?

I am about to start putting my new build together, and I came across a blog where someone used the same case as me. They had an issue where the sata ports on the motherboard were angled sideways which made it difficult to plug in the sata cables due to being really close to the wall of the case. This is how tight it was for them: https://lanoc.org/review/pc-hardware/8346-project-data-hoarder-truenas-core-build-part-4-psu-rails-and-installation?start=2 My motherboard also has sideways facing sata ports and the sata ports are in the same location on the board, so I think I will have the same problem - especially since the cables I will be using look like they have bigger ends than those used in the blog. So, I am thinking this could be solved with some angled adapters or similar. I have found some online such as this: However, since the sata ports are stacked one on top of the other, I would only be able to use one of these for each column of ports (4 in total). I was thinking I could get 4 short adapters and 4 long adapters (which reach over the short ones), but I haven't been able to find any long ones. Does anyone have any other ideas for solutions to this problem?

Yep good points. Thanks for the suggestions everyone!

This is really cool. Great find. Makes the decision easier knowing this is an option. I think that makes sense. I can't see myself needing more than 96GB and starting with 2x16GB is a cheaper entry point and will give me time to see what my RAM needs really are and how it grows before jumping up to 32GB modules if need be. Just looked it up, it looks pretty decent. Can only find it in one shop here but it is a bit cheaper than the Corsair and Silverstone which makes it one to consider. Thanks for the tip!

Thanks for linking those posts, interesting reading (with the help of google translate 😋). Also I saw you linked in one of those posts you have the same motherboard and the i3 8100. I see why you liked my choices Wow yeah 50% or lower would not be good. Good to see these two PSUs are much better than that! Hmm ok that's interesting. Opens up options a bit. That Kingston module seems to be pretty available locally (Australia), although it is about $120 more expensive, I guess because of the higher density and higher speed. I think the Samsung M391A4G43MB1-CTD would be good but I can't find it for sale in Australia. I did find some Crucial ones though - 3200MHz and 2666MHz (that one says REG in the title but I'm pretty sure it's not, according to Crucial's website). Now I just need to convince myself it's worth paying a little extra to get expandability to 128GB in case I need it one day. And then decide between 3200MHz and 2666MHz.

Yeah good point. 6 cores would be nice but I was thinking it might be overkill for my applications at the moment - using more power and costing more initially. I thought ECC memory would be a decent upgrade as people often recommend it for servers and I like the sound of better stability, but I guess it might also be overkill for a small server such as this. Hmm this is interesting.... I just had a look at other cables on that store. Would this one be the correct cable? They look the same but look to be in opposite directions. However the one I picked has "Re" in the name - does that not mean reverse? Also the image on the page in the one I picked has SFF-8087 side going to the HDD backplane and the SATA 7-pin going to the motherboard sata ports. I thought that's how it would work, since my backplane has an 8087 port and the motherboard has sata ports? Why doesn't that work? Sorry if these are stupid questions, I'm new to SAS.

I've looked up some data on the Corsair, and it does seem to be really efficient at low power. However comparing to the Silverstone I think it seems pretty similar? (Not trying to prove you wrong, or defend the Silverstone - just trying to make sure I'm not missing anything). https://www.corsair.com/corsairmedia/sys_master/productcontent/WW_RMx_Series_2021_QSG_AA.pdf https://www.anandtech.com/show/10207/the-silverstone-strider-platinum-550w-psu-review/4 Corsair RM550x: https://www.tomshardware.com/reviews/corsair-rm550x-power-supply,4484-5.html Silverstone 550w: https://pcper.com/2016/10/silverstone-strider-platinum-550w-power-supply-review/5/#ftoc-heading-6 Seems like at 55w they are both around 84-88% efficient and at 20w they are around 75-78% efficient. It's kind of hard to tell exactly without precise markings on the graphs, but they seem pretty close. Do you have any other links about the Corsair's low power efficiency vs others? Or is the Corsair just better at going down even below 20W which isn't measured by these charts or something? Yeah I agree 32GB is plenty for the time being. I was originally looking for a single 32GB stick to leave upgrade options open (and saving power would be nice too), but I couldn't find any 32GB sticks at 2400MHz that were unbuffered - they all seemed to be buffered/registered. Like you said - availability really limits choices.

Thanks so much for taking the time on this. I'll look into the Corsair RMx 550. I assumed the platinum rating on the Silverstone would make it more efficient than a gold one but I guess not necessarily... It looks like the Corsair is a bit longer which puts it at the 160mm stated limit of the case. It should still fit, but you're right - it might be tight. Sorry, that 2.5 HDD is actually 3.5. I must have selected the wrong one in PC Part Picker and didn't notice. I thought the CPU only supported 64GB RAM so that's why I was planning for expandability up to 64GB. This is from the Intel spec sheet: Unless that doesn't mean what I think it does?

Hi, I was hoping to get some more experienced eyes to look over my planned build and let me know if you see any problems or have suggestions. I'm upgrading from a really old store-bought desktop PC with a dual core Celeron CPU and 8GB of RAM, that I threw some hard drives in. I have found that my RAM is often completely full which causes problems for my dockers, and the CPU maxes out as well. The server is used for file and media storage, backups, standard media apps (Sonarr/Radarr etc.), Nextcloud, and a bunch of other dockers eg. Bookstack, Paperless, Linkace, Unifi Controller, Calibre. Note - there is no Plex/Jellyfin as my media server runs on my Nvidia Shield so no transcoding is needed. In the future I will probably keep adding dockers when I find fun new projects (probably Home Assistant at some point). No VMs at the moment but I'd like to have enough power to be able to run one if I ever have the need. Here are the parts I am looking at: PCPartPicker Part List: https://au.pcpartpicker.com/list/gXJ7Dc CPU: Intel Core i3-9100 3.6 GHz Quad-Core Processor Motherboard: Gigabyte C246M-WU4 Micro ATX LGA1151 Motherboard Memory: Timetec Hynix IC 32GB KIT (2x16GB) DDR4 2400MHz PC4-19200 Unbuffered ECC 1.2V CL17 2Rx8 Dual Rank 288 Pin UDIMM. Note: will look to expand to 64GB eventually if 32GB is not enough. Cache: 2 x Samsung 970 Evo Plus 500 GB M.2-2280 NVME Solid State Drive (mirrored) Storage: Seagate BarraCuda 4 TB 2.5" 5400RPM Internal Hard Drive Storage: 2 x Hitachi Deskstar NAS 8 TB 3.5" 7200RPM Internal Hard Drive Storage: Seagate IronWolf 8 TB 3.5" 7200RPM Internal Hard Drive Storage: 2 x Seagate IronWolf 12 TB 3.5" 7200RPM Internal Hard Drive Power Supply: Silverstone Strider Platinum 550 W 80+ Platinum Certified Fully Modular ATX Power Supply Case: SilverStone RM21-308 Black 2U Rackmount Case Cables: 2 x SilverStone CPS03-RE 50cm Mini SFF-8087 to SAS/SATA Cable Case Accessory: SilverStone Sliding Rail Kit for SilverStone RM21-308 Chassis Questions: Do you think the CPU is a good choice and powerful enough? I picked it because it supported ECC RAM, has integrated graphics (mostly for BIOS, but also just in case I need it) and Xeon CPUs seemed really expensive. Would have gone newer but 10th gen i3's don't support ECC. Any thoughts on the motherboard and RAM? That RAM kit is one of the only ones I could find with 32GB unbuffered ECC 2400MHz. Does it look like I've forgotten any parts? Cables I will need etc? I've never used hot swap bays or HDD backplanes so not sure if I've got everything to connect the hard drives. Notes: No budget really - just want something powerful enough, but not overkill. I already have the hard drives.

Ok that's great. Thanks very much!

Good question...So in the process of getting the server out, plugging in a screen and trying to boot into the BIOS the cache drive seems to have come back online in the GUI. I never managed to get into the BIOS because I think I needed a VGA cable instead of DVI but that's another story. While I was in there I noticed the cache drive was plugged into a little sata expansion card, so I plugged it straight into the motherboard in case the expansion card had something to do with it. But everything is back online now - cache shares, docker etc. all seem fine. However there is a SMART error on the cache drive. How worried should I be about that? When I click on it it shows the reallocated sector count is 1. Is that like a "keep an eye on it" kind of error or a "stop using this drive now" kind of error? New diagnostics attached. tower-diagnostics-20211014-2016.zip

Hi, last night my cache drive disappeared off my server. It was working fine during the day and the server hasn't physically been touched or moved in weeks. I restarted the server but the cache drive didn't come back after reboot. It's not appearing in the main page, dockers are gone, and shares that use only the cache drive are gone. Can anyone help me work out what the issue is with the cache drive? Has the drive gone bad and should I throw it out, or is it still usable? I have attached my diagnostics after the reboot. I also have diagnostics from before the reboot and can upload if that would be helpful. Thanks in advance. tower-diagnostics-20211014-0925.zip

Thanks for that. Cache didn't come back after reboot so I will make a new thread.

Hi, I am getting an error message when I go to the CA app store that tells me to post here. It says: I also have error messages on the docker tab: There is one of those for each docker installed and also all my docker containers aren't working. I assume it is related as both errors mention line 331. Any idea what's gone wrong? tower-diagnostics-20211014-0000.zip

Hi, I want to install paperless-ng but it appears the template is no longer in the CA App store? When I search for paperless I only see the old paperless, and when I search paperless-ng I get nothing. I can see it if I click "get more results from DockerHub" but then the template has no pre-filled variables and paths. Has it been removed for a reason? Or am I doing something wrong?

I didn't figure it out but it seems to be working a lot better now. Maybe I was wrong or just confused about it overwriting itself, I'm not sure. But I'm not having any problems with it all now.

Thanks for that @Richamc01, I've managed to get it working after a bit of messing around. Looking at your config made me realize I forgot to change the server_name from bookstack.* to janddybookstack.* to match my duckdns subdomain (it's not janddybookstack, but you get the idea). Once I made that change the reverse proxy was working, but the page wasn't loading correctly. I remembered reading something about that before and sure enough, I scrolled up and @klogg had the answer: Removing https:// from the APP_URL in the .env file fixes the problem, but restarting the container seems to overwrite the manual change and puts the https:// back in. Weirdly, even after the https:// is back in the file, loading the page still works. However It didn't work when I tried it on my phone at first, but then I removed the https:// again and it worked on my phone from then on. Maybe it's cached in each device's browser or something? Not a huge deal, but if anyone knows how to stop the .env file from being overwritten on restart that would be cool. Anyway everything is 99% functioning now so thanks @Richamc01 and @klogg for the assist and thanks to the linuxserver team for the docker

That's good to hear. I think I only have my own subdomain - something similar to janddybookstack.duckdns.org. Your config file would be great if you don't mind.

Is it possible to use duckdns for reverse proxy with bookstack? I followed gilbn's guide but when I got up to the Nginx section it didn't seem to work. I noticed in the comments of the code that I had to copy into the 'default' file in the swag appdata that it said Does that mean the entire reverse proxy won't work, or just that specific feature won't work? Sorry I'm not very familiar with web stuff like Content-Security-Policy. I have swag reverse proxy set up for multiple other containers using duckdns and I have also have bookstack running fine locally. I understand if the solution is to just buy a domain but I wanted to see if it was possible to do it the way I have everything else set up first.

I finally managed to figure out what was causing this issue. My PIA password included the apostrophe symbol: ' which I guess was interfering with the docker code somehow - maybe ending the openvpn file early based on the "unexpected end of file" in the log. Changed my PIA password and all was good. It was odd though that I've been using this password for a few years and it only started causing problems recently. Oh well, just glad it's fixed. Hopefully this helps someone if they have the same issue.

Hi, since about a week ago my deluge docker has stopped working and I'm not sure how to fix it. When I start the docker, the logs have these errors: 2020-10-03 13:44:13,427 DEBG 'start-script' stderr output: /root/openvpn.sh: eval: line 73: unexpected EOF while looking for matching `'' /root/openvpn.sh: eval: line 74: syntax error: unexpected end of file After a minute the docker seems to be using 100% of my CPU and RAM, other dockers start crashing and the Unraid web-ui goes non-responsive. I have tried removing the docker (and also removing the image) and reinstalling. I also tried rolling back to an older version from 2 months ago which gave me a different error: 2020-10-03 14:03:00,929 DEBG 'start-script' stdout output: [warn] Exit code '52' from curl != 0 or no response body received [info] 7 retries left [info] Retrying in 10 secs... I also followed the instructions to upgrade to PIA Nextgen thinking that might be the problem, but no change. Does anybody know what the problem is and how I can fix it? Thanks in advance

Yes appdata is listed on the shares tab and I can browse to it in windows explorer. The default in Settings - docker is /mnt/user/appdata/ so I'll change the source to that when I set back up.

Ok thanks, the source must exist then because the backup ran this morning. Does that mean there is an issue with my source setting, since it backed up nothing? Should it have been /mnt/user/appdata/ instead of /mnt/cache/appdata/?