FlamongOle

I found the issue, it was a certificate problem. Restoring the previous one fixed it. nginx -t revealed it, thanks!

After upgrading the server today, the WebGUI refuses to load. It looks like it fails with the GraphQL-thingy. I don't use that myself, but had it enabled with a plugin I tried out but later disabled it. I don't know exactly what went wrong, but it's not sure it's related to the Unraid upgrade.

Upgraded to 7.3.4 from 7.3.3, server starts but I can't access webgui anymore, and logs are filled up with: Oct 26 17:05:53 Odin webgui: curl error: Failed to connect to localhost:80 over unix:///var/run/nginx.socket after 0 ms: Could not connect to server endpoint: diagnostics Oct 26 17:05:53 Odin publish: curl to diagnostics failed

Update 2026.10.12 Commit #528 - MINOR: zfs_disk() substring match: sda matched to ZFS member sdaa1. Thanks to kapsaicin@github for the fix.

Set the LED signals to "Static"

I will check the comment fields later. Thanks for reporting.

Use -4 instead of 4, negative numbers allowed

Update 2026.10.02 Commit #524 - FEATURE: Added cold storage option to save information of a drive not attached or online. Commit #519 - MINOR: Make the tray map usable on mobile viewports (Thanks to joshmans@github). Commit #518 - ENHANCEMENT: Fix pkill quoting, add JSON export, add PHP lint CI, harden shell_exec() calls (Thanks to joshmans@github). @getgray you can try this update, I have not been able to test this myself other than that it won't destroy anything and that it does store the set values. To use cold storage: attach the drives to your hardware do a force smart+db update go to Tray Allocations assign the drives how you want them check "cold" for those that's gonna remain as cold storage save pull the drives marked as "cold" from the hardware test by either force smart+db or reboot.

Important information! There's a new plugin emerging from a fork of Disk Location, it's apparently vibe coded from what I can see and it's called "Disk Location Next" I am not affiliated with that plugin what so ever, and I will request that user to rename the plugin so it won't look like it's affiliated to my work. You, users, are free to choose whichever direction you want, maybe the new plugin will be better, or even replace this one in the future. I am very happy that someone will provide with an alternative, choices are good. I will still maintain and continue to build this plugin the way I want for now, without the help of an AI. Call me old fashioned, but I am not vibe coding my plugin for several reasons, the most important is: personal lack of control and that it is very bad for our planet. Update 2026-10-18 (ISO date): "Disk Location Next" has been renamed to "Bay Watch", and will likely show up in the Apps soon. Check out the project at: https://github.com/joshmans/unraid-baywatch

Hello getgray, Drives will be automatically removed/moved to history when it's not detected by the system anymore. This happens during boot of the system or during Force SMART+DB. Things might have worked differently in the past, and was most likely a bug that shouldn't work that way. For you it was a feature =D Adding a possibility to add a cold storage is a good idea, and I will list it up as a feature request. I will have no ETA on when this will be implemented. However, this will require the drive to be added and scanned, then marked as cold storage before it will be removed. If it's already in the history tab, the location of the drive is of course removed as well. It should however be relatively easy to implement (but other things might break .. :p). Technically is all about adding a status letter, but it has to be acknowledged by rest of the plugin.

I get these errors as well: Stack overflow in unmanaged: IP: 0x14dfe318590a, fault addr: 0x555a4ed4699b But I have since start and I'm been using it for over a year. Despite the message, the server starts anyway. I never really cared about it and dunno if it matters I'm running two servers from two separate docker containers.

Set group name alignment: hide

You can add bypassed/hidden drives: so 2*5 instead of 2*4 in layout, then hide the empty slots at the right side (checkmark): You can also just create a hidden layout you put between each section. 1*2, hide both drives and the name:

Can be if the drive was disconnected or so during a reboot. Disk Location moves not found drives as "lost", if it's detected again, it can be assigned.

Update 2026.08.21 Commit #515 - MINOR: Adjusted dev_by_phy_path to include partitions that exists to include more drives to be physically attached in case non part-names does not exists. This update should have minor/none effect on most systems, but might include more devices to be stored as "physical"

Well, is it assigned at all?

It shows tray numbering from the largest group. Setting it incorrectly will fail the "save". Use the numbers that makes sense, don't try to force it off the group - it won't work.

Sorry for the late reply, forgot to follow the topic. No, this is what I see:

Update 2026.07.09 Commit #513 - BUG: SMART+DB fatals when by-path is non-existing. Some systems do not have /dev/disk/by-path apparently, so the SMART+DB update would fail. Checking if the directory exists before checking the contents with dev_by_phy_path() function.

That makes sense. Expected that this path would always be there. I'll make a fix soon

Cannot do anything with that information. Please send me Disk Location debug file and also PHP logs from Unraid. Also share with me if you did something else before the problems happened. In System tab: Enable debug, then press Force SMART+DB, cancel it if it doesn't do anything withing 5 seconds, disable debug, then send me the file. Or uninstall it.

Update 2026.07.08 Commit #511 - BUG: Physical assignments wont move with the groups. Added physical.json into the rewrite loop with locations.json.

Update 2026.07.07 - Physical tray assignments has arrived! Commit #506 - FEATURE: Physical assignment of trays. Some devices can be used to assign a tray to a physical address. Using this will automatically replace drives added to that tray. Device path are assigned to trays instead of device name and serial number. READ THIS BEFORE USE OF THE NEW FUNCTION: Some devices can be used to assign a tray to a physical address. Using this will automatically replace drives added to that tray. Device path are assigned to trays instead of device name and serial number. This requires a drive to be attached and assigned before it can be used. IMPORTANT! If there is a system change with e.g. HBA/SAS card (even motherboard PCI-slot), SAS-SATA breakout cables or motherboard, the physical locations must be wiped and reconfigured in the plugin. Using HBA/SAS cards in a fixed PCI location should normally use the same addresses. If such a major change is required and you want to preserve the tray locations: Press "Save" on "Tray Allocations" page to update the manual location database. go to the "System" tab and create a backup. then wipe ONLY the "physical" configuration BEFORE doing the changes. Hot swapping requires execution of Force SMART+DB, and will be prompted within 5 minutes. Inserting devices in powered off mode, will automatically execute it during boot. This text is also found under "Help" section in the plugin. ADDITIONAL INFO: This function is optional, if you don't want to opt in, you never have to. To assign physical locations, each location must be assigned and "physical" checkbox must be marked per drive. Drives not compatible will be disabled (e.g. NVME drives). POSSIBLE FUTURE PHYSICAL LOCATION TEMPLATES: It might be possible to make location templates from original server systems, like the 45 Drives Unraid system but also others. To create such templates for me, I would need this from you, everything below is mandatory and must be completed as written: Information about the server/enclosure name and model including HBA/SAS card name and model. All trays must be filled with drives and recognized in the plugin. All tray locations must be saved how it physically looks like on the server. Physical checkbox must be marked and saved. Then send me this file: /boot/config/plugins/disklocation/physical.json DO NOT SEND ME FILES FROM NON-ORIGINAL CONFIGURATIONS!!!

Should work now, forgot to increase the version number in the install file. After "Force SMART+DB", you can manually spin down drives, then run just "SMART" afterwards to see if it behaves as expected.

Update 2026.07.06 Commit #507 - BUG: SMART data gets saved even if it's empty. SMART data got pulled when there's nothing to fetch. Checks if SMART is available and enabled before saving new SMART json files. @Lean_ try this update. You must do a SMART+DB to fix the database before we know.