FlamongOle

@grupoleon Can you give me the entire output of smartctl -n standby /dev/sdb

I see what's going on now. It's indeed caused by the 16:0:0:0 virtual device. The problem is that smartctl recognize it as "ACTIVE", then it is cross-checked with the array of missing devices. If it find uneven amount of devices (those with serial numbers and those without) it will post that in the dashboard to warn about incomplete smart data fetch. I need to find a way to ignore the false devices, likely force them to be "UNKNOWN" during powermode check.

And if you click on "Dashboard" tab now?

I am mostly curious of what 16:0:0:0 (possibly still "sdb") is. I'm not sure if it's related, but it the only thing that doesn't make any sense here for me. It's detected as ACTIVE, but yet give no serial number. What happens if you do a lsscsi -bg | grep "16:0:0:0" Insert /dev/sdX from above, eg.: smartctl -x --all /dev/sdb

Nah, that script is an entire separate thing. What do you get if you type: ls /tmp/disklocation/smart/and cat /tmp/disklocation/powermode.json

Update 2025.05.27 Commit #432 - BUG: Further enhancements for checking arrays and variables. @grupoleon try this update, at least it should get rid of the PHP errors above. You might need to reset "devices" and do Force SMART+DB afterwards.

Any disk location related errors in the PHP log?

Update 2025.05.26b Commit #440 - BUG: Devices with no serial numbers can create an empty array and cause the plugin to crash. Made adjustments into cronjob, deleting "devices" and running Force SMART+DB required afterwards. Commit #438 - MINOR: htmlspecialchars duplicate where it tries to format itself in the group names. Removed htmlspecialchars from the variable to be stored in json array, it's not needed as it is read by htmlspecialchars. Commit #437 - BUG: Assigning not working and deletes all assigned devices if tray numbers are hidden. Added bypassed variables to the array instead of variables that were unset. @grupoleon you can try this. You must likely delete "devices" then run Force SMART+DB afterwards. You should be able to access it via "Disk Location - Recovery" page under Tools. Go to "Reset plugin", select "devices" then "Reset", then do Force SMART+DB.

From what I can see in the logs, disk3 and disk22 is listed but does not have serial number and name, and thus will be missing an ID. Do you know what these are? Is it a controller of some sorts, is it a dead one, CD-drives, is it mapped drives from proxmox? I have a CD-drive in mine as well, which causes no troubles. I can probably fix this issue with ignoring empty inputs (which I though I already had). But would be nice to know if you know what this is.

Update 2025.05.26 Commit #438 - MINOR: htmlspecialchars duplicate where it tries to format itself in the group names. Removed htmlspecialchars from the variable to be stored in json array, it's not needed as it is read by htmlspecialchars. Commit #437 - BUG: Assigning not working and deletes all assigned devices if tray numbers are hidden. Added bypassed variables to the array instead of variables that were unset. @devalias should be fixed now :)

Haha, whoops! This likely happens that you have hidden numbers on the trays, and why underneath it can't connect TrayID with Physical tray number :P I didn't have that in my mind, but it should probably be easy to fix. The 2.5" problem seems to be catching on a htmlspecialchars/htmlentities loop where it tries to format itself. This should be easy enough to fix as well. But I believe the name will extend every time you save it with additonal "amp;".

Update 2025.05.25 Commit #435 - MINOR: Saving tray allocations with unassigned devices throws error messages. Added a check if the variables are empty or not. If the value is empty, then it will be ignored, otherwise throw error for specific configuration.

Ah! Thanks for that info, now I remember why this was earlier ignored to post an error message :P I'll remove that again unless I find a way to distinguish that.

Update 2025.05.24 - Commit #433 - MINOR: Unknown USB bridge alert. Alert message popped up on devices that shouldn't be marked as failed. Added another variable check to prevent false reports. - Commit #432 - BUG: Multiple users reported blank screen when saving drives to layouts, checking if its an array, otherwise set variable to null. Haven't had time to test it properly, but please give it a go. @noonamer @zdzosel @awestera

Yes, just ignore this. I thought this was already fixed, looks like I need to look a bit further into this

Please give me the relevant output of the php log. Do you create stuff from scratch, or do you expand current setup?

Ye, don't assume things just by looking at the benchmarking. It's rather more relevant for the mechanical drives anyway and it's rather an indicator if a drastically speed decrease happens over time. I would read too much into it with SSDs.

Update 2025.05.22 Commit #430 - IMPROVEMENT: Assignments are now done with physical tray layout, and not TrayID (it's still TrayID under the hood, but hidden from GUI). Commit #429 - IMPROVEMENT: Added possibility to reset temporary files. Reset "temporary files" under System tab will now delete SMART data and others in tmp. Force SMART required after execution. Commit #428 - ENHANCEMENT: Fetch more data from SMART files rather than DB. SMART files in temp folders holds more recent data, so this will be prioritized when reading increasing parameters. Commit #427 - MINOR: Benchmark for removed devices visible. Hide benchmark for removed devices, these can still be retrieved by raw download. @Frank Ruhl and @CheekyPanda (and probably others), you should be happy now (#430, if it works :p)

It's been answered here already. Please update to latest version and run Force SMART+DB.

Update 2025.05.19 Commit #425 - BUG: Flash devices etc. alerted falsely. Added more SMART checks to prevent this from happening. FORCE SMART+DB update must be done for those affected with this bug.

no.

You can ignore this error for now, I'll find a way to skip memsticks/Flash next update.

Can be several things. If the drive is in use while benchmarking, the values will be lower. But likely this is the actual speed of the SSD, cache is bypassed (which you'll often see higher speeds from). Some SSDs are in fact just slow in read and write, but they still feel faster than spinning drives because of no seek time.

Likely how it's connected or that the controller for it does not pass through Smart or something. Usually Flash is never even included in Disk Location as it's mostly just pen drives. I have no plans to implement it either.

Update 2025.05.17 Yes I know, one day too early but who cares? 😛 Commit #423 - MINOR: Other irrelevant errors occurred in SMART tab. Rearranged the code. Commit #422 - ENHANCEMENT: Added notification when a major device error occurs. E.g. If the drive loses it's controller, send an alert notification in Unraid about it. @Midas_Tech this version, then a SMART update should hopefully fix it! Happy Norwegian Constitution Day release! 🇳🇴🥳