Mattaton

Finally got this sorted. Changed all the keys in the template to match the repo and then added DB01_AUTHDB key for the postgresql backup to work. The postgresql dump wants there to be a db with the same name as the user, so the extra key is needed.

My problem, in this case, ended up being a bone-headed mistake on my part. After upgrading my server hardware, I used the old mobo as a desktop PC. Forgot to update my pfsense router so that the same static IP wasn't assigned to the mobo's NIC MAC. So when I fired up the new PC, I had an IP conflict and that threw my server out of whack. Still though, the widespread issues with the i226 NIC bothers me. Two different Z790 mobos in my new desktop had the same issues. I'm assuming a linux-based system could have the same problems. I have a new discreet NIC coming for the server now. I'll swap it out to avoid any potential issues.

Last night my unRAID server started acting flaky. First noticed that I couldn't access Plex properly. It would sometimes load content and then not. Then I tried pulling up the unRAID webgui. It was slow moving between pages. I checked docker containers like cloudflared and it was spitting out connection errors like crazy and obviously any sites I was using a CF tunnel on was intermittent. At first I thought there must be a cloudflare issue, but that shouldn't have affected Plex, so it seems the culprit must be network connectivity based. I recently rebuilt my server with an ASUS z790 mobo with the Intel 2.5G i226 nic. I also built a new desktop with a similar mobo and the same nic. No end of issues with it on Windows until I finally disable the on-board nic and installed a discreet one. So, I'm wondering...is unRAID just as susceptible to the i226 issues? I've been running this build for about a month and it just started doing this. I have a new nic on the way, just in case. I'm hoping that will solve the problem. I'm at work right now and I can't access the server at all via VPN like I normally can. I started a reboot this morning before I left and it either never came back up or the network connection is completely gone now.

You can put the images anywhere you want. Anything you can browse to under shares (at least in how I tested it). I just tested this... I found an image in a share. I clicked on the image and it opened/previewed over the unRAID interface. I right-clicked the image and chose Open in new tab. Then copy the url of the image tab and paste it into the folder settings.

Cool, yeah, the selectors when trying to "hack" an existing layout after-the-fact is always fun. Trying to override existing styles with tighter specificity and all that good stuff. 🙂 I haven't looked yet, but I may already have selectors for these items (or close to them) that don't need has()....maybe. 😄 The styles I have on github do include the dashboard. I just have it all split up into several files so users can easily find the area they want to tweak or drop altogether. Thank goodness the CSS folks introduced custom properties. So nice for this sort of thing. Thanks, bud! Now, go get yourself some sleep! 😄

Noice. Thanks! Would border-radius: 50% not work for some reason? That should make a circle even if the user has adjusted the icon size. I was going to look at this today, but last night my server decided it wanted to misbehave. Can't even access it at all this morning. So, needless to say, my plans have changed. 😄 I'll incorporate this as an option in my themes when I get a chance.

You can do almost anything you want visually. I have some fairly extensive and customizable themes on GitHub. I don't think I have anything in there for background color and I'm not sure how we'd do a circle around it, but I can try to help you with it. I'll try to have a look at it tomorrow and see what I can come up with.

I tried that earlier and the VPN manager page just went blank. I had to go into the etc/wireguard folder and delete the wg0 files to get it to come back.

I am trying to google this, but everything I see is talking about setting up a VPN on unRAID, connecting to unRAID via VPN, not having unRAID connect to a remote network via VPN. I have Wireguard configured on pfSense in my home. I have an unRAID server in a remote location that I want to have "dial in" to my home network and basically appear as though it's on my local network. I'm looking at the VPN Manager, but as a relative novice--especially with Wireguard--I'm having a hard time equating the terms to what I set up for a Windows machine as the "client." (I know Wireguard doesn't use the terms server and client and that basically what I am considering the server and client both have tunnels and peers set up from their POV. - That just makes it more confusing when unRAID's VPN settings are using slightly different terms.) So, how do I only set the unRAID Wireguard to be the "client" peer to my pfSense Wireguard? Looking at the Wireguard Quickstart, but the scenarios are a little off from what I'm trying to do. LAN to LAN and VPN Tunneled Access are close, but not quite. I don't need access to the LAN on the unRAID server side (only need access to the server itself, along with any containers it may be running) and I don't want all the unRAID server's traffic going through the tunnel. Thanks!

I ended up with G2s, so hopefully I'm good 🙂

I think what was meant (or at least how I took it) is that if I potentially bought a used one on eBay, the GUID may have already been registered with unRAID and therefore would be worthless to me. I get how it works and what the issues are, which is why I rely on folks like @Hoopster for some real-world experience with a particular device. Unless he is the only one ever successful in using one of these (which I suppose is possible - but it doesn't sound like it), then it probably has a unique GUID per reader, and not per model. I certainly appreciate the caution! Thanks! 🙂

They're supposedly brand new, in-box. So I hope this won't be an issue for me. But yes, you're right. Thanks!

Follow-up to my previous post. I tried changing the key names in the template to better match what's shown on the db-backup github page. That got rid of some of the deprecation warnings, but it still mishandles and shuffles the values when the script runs as in my first attempts. From what I see on the github page, it also seems like you can back up more than one type of database (or multiple instances of the same type) from one container instance by prefixing the key names with DB01, DB02, etc. It'd be great if this functionality could carry through with the unRAID template. I wonder if this template was initially only meant to work with MySQL/MariaDB and therefore never tested with PostgreSQL. @A75G I hate to be a bother, but if you could offer some insight, I'd love to get this working. Thanks!

Nice. I was able to find one on eBay. Looking forward to giving this a shot. 🙂 P.S.: "Vacation Home"....I'm jelly 😄

Awesome! Glad you got it working for you! And hey, we've all been there. Thanks for being cool about it. 😊

Ah man. That's awesome, but a shame they aren't available any longer. Any suggestions on alternatives?

Such is the nature of a community-driven plugin ecosystem. Which isn't unique to unRAID. unRAID didn't create the old plugin or this one, so there would be no warning from them. The development for the old plugin died and I'm glad @scolcipitato created this one. He's very responsive and helpful in the development and improvement of it. Can you elaborate on what isn't working for you? We'll be glad to try to help you with it. I think most of us have FolderView working just as well or better than Docker Folders at this point.

Slightly off-topic, but is there any truth or evidence to the claims where people have used SD cards in USB readers/adapters or SSDs connected with USB cables? Supposedly unRAID pulls the ID of the reader or SSD holder/connector and registers that which then allows the user to swap the SD card or SSD if it goes bad. So as long as the reader or holder/connector is good, then unRAID is "happy." If this indeed works, it sure takes a lot of the headache out of swapping flash drives. Just pop out the bad SD card, restore to a new one and you're good to go. I haven't tried it, but I do have an external enclosure for NVMe SSDs. When I connect it to my Windows machine, it sees the enclosure as the device and not the SSD. Maybe that's how unRAID would see it, too? If it has a GUID, you're golden? When I connect the enclosure, it gives Vendor, Product, and Serial Number. Although the serial number doesn't seem quite unique: 012345678930 So, yeah, probably not with this. I just plugged an SD card into a little USB reader/adapter. The reader comes up with a more unique serial, but still nowhere near the length of the serial on a flash drive. Vendor ID : 0x0BDA (Realtek Semiconductor Corp.) Product ID : 0x0120 Serial: 200604130921XXXXX

I had read the first PSA post of that thread, but definitely not the entire thread. So, I knew to steer clear of Sandisk...although I still find a lot of rather current posts on Reddit and elsewhere that people are using SanDisk with no issues. So I assume this comes back to the counterfeit crud on Amazon. Maybe they got them at a Big Box Store or something instead of Amazon. I've never heard of Kanguru. I guess if they're tested. I know the flash drive doesn't matter THAT much as long as it works. When shopping, reputable brands just give better odds of working, I suppose.

Are you guys having any luck finding decent USB drives on Amazon that are proven to work for unRAID? I have bought Samsung BAR Plus and FIT Plus in the past based on SpaceInvaderOne's video, but even those aren't available on Amazon anymore. They're 64GB and up. Which makes sense from a typical use-case for the drives, but not so much for unRAID. And those weren't even USB2.0. That's even harder to find. And with Sandisk no longer being recommended, that cuts out a big reputable brand...although not so reputable where unRAID is concerned, I suppose. Where are you guys landing drives these days? About to get a license for a new server and I need a drive! 🙂 Thanks!

I need some help with the db-backup container... Backups of MySQL server in separate container work fine. Instance for Postgres fails. Postgres logs say: 2023-11-21 03:00:00.852 EST [535] FATAL: password authentication failed for user "dbbackup" 2023-11-21 03:00:00.852 EST [535] DETAIL: Role "dbbackup" does not exist. Connection matched pg_hba.conf line 100: "host all all all scram-sha-256" 2023-11-21 03:00:00.872 EST [536] FATAL: database "admin" does not exist But DB_NAME and DB_USER are not set to be admin or dbbackup. So where is it getting that from? The login credentials the container has are correct and work fine to log in via Adminer. If I run backup-now in the container terminal, I get this: psql: error: connection to server at "192.168.75.10", port 5432 failed: FATAL: database "admin" does not exist Are the keys in the template wrong for postgreSQL? Manual run with debug mode on shows this as the command sent: psql -h 192.168.75.10 -U admin -p 5432 -d admin -c 'COPY (SELECT datname FROM pg_database WHERE datistemplate = false) TO STDOUT;' Looks like it's set the port number as the password (unless -p is for port?) and the username (admin) as the database name? Here are my settings. It looks correct, but it doesn't seem the variables are being parsed correctly for postgresql. Tried it all on a different unRAID server. Same error. It comes back saying database "-DB_USER-" does not exist. There is something broken in how the query is parsed from the container variables. They don't work the same from mysql to pgsql. Other than the above issues...I am also seeing a lot of deprecation warnings and the keys for the template don't match the keys spec'ed on the github page for db-backup. 2023-11-22.09:02:49 [WARN] ** [db-backup01-scheduler] Deprecated and unsupported variable 'DB_DUMP_BEGIN' detected - Please upgrade your variables as they will be removed in version 4.3.0 2023-11-22.09:02:49 [WARN] ** [db-backup01-scheduler] Deprecated and unsupported variable 'DB_DUMP_FREQ' detected - Please upgrade your variables as they will be removed in version 4.3.0 2023-11-22.09:02:49 [WARN] ** [db-backup01-scheduler] Legacy unsupported variable 'COMPRESSION' detected - Please upgrade your variables as they will be removed in version 4.3.0 2023-11-22.09:02:49 [WARN] ** [db-backup01-scheduler] Legacy unsupported variable 'SPLIT_DB' detected - Please upgrade your variables as they will be removed in version 4.3.0

I changed the Tools->PHP Settings->Error reporting level to Errors Only (it was set to Default) and that got rid of the warning and brought the UPS details back.

I just noticed an error (PHP Warning) on the UPS Settings page. Under the UPS Details portion where it used to show, well, the UPS details, it now just says this: Warning: Undefined variable $freq in /usr/local/emhttp/plugins/dynamix.apcupsd/include/UPSstatus.php on line 90 Any ideas how I can fix this?

Ah man. Sorry. Maybe my new hardware/build fixed this issue, but I still had the issues in the other thread which ipvlan ultimately fixed. Thought the two problems were related, but maybe not. Good luck with it!

My issue ended up being that I was using macvlan for custom docker networks instead of ipvlan. See this thread: Hopefully that helps you!