FlamongOle

line 95 makes way more sense than line 9. You do not need to worry about that, it is only a converter for the harddisk size to convert from pure bytes to eg. TB. It faults when it does not receive anything. I'll fix this later.

Update 2019.02.18 Commit #53 - MINOR ISSUE: Replaced single quote to double quote for the definitions. Could cause some warnings to appear - maybe.

I might have fixed the warning message, but I'm not 100% certain about it. Next release will maybe help. It looks like the SMART data can't be read from the last SSD drive you have, as you can also see from the unraid page. Run smartctl manually on the drive and compare it with the other one. Are they connected on the same board? Maybe it might be some hardware controller issues, sometimes (rather often) a motherboard have multiple types of chipset/sata controllers on board. I might look into a mini version, but it has to be cached and running in the background in that case. As it will read smartdata to have any actual purpose, it might take time to do so and also interrupt the system drive by drive until it is finished. I dunno if that would be a good solution. I'll add it at github for possible future tasks, but don't count on it (at least for a while).

Yes, it is intended to work like this as people might change entire configuration and do more than one step at a time. Reorganizing of drives, adding or removing and then also swapping controller or motherboard would be hard to keep track on. That's why it is chosen to just use harddisk+slot.

Update 2019.02.14 Commit #51 - FEATURE: Added warranty left and comment options to be visible in the devices tab. These options has to checked and be saved in the configuration. For existing users, both comment and warranty must be activated and saved manually. For new users the warranty left is on by default.

Did it work without issues?

1: Yup, and believe me I wanted to avoid this as much as possible - and luckily won't be any problem in 6.7+ The first official stable version of this plugin will be near Unraid 6.7 stable launch, and then it will be required to use 6.7+. My main focus is to keep it working with the current stable release, and I like the current version number so no rush. 😈 2: I believe the plugin install script would reinstall them as the plugin itself? The packages and the plugin is stored at /boot, so those should "survive" (reinstall) a(t) reboot?!

Update 2019.02.11 Added max version to the package installer so it won't downgrade the package by Slackware's logic of versioning for Unraid 6.7. nothing else 😛

I totally agree and will add this soon, the lack of documentation is horrible though. Also any proper distro have a dependency check and don't mess with "different version number" but higher and lower, I really wonder why it was even settled for Slackware as a base of all distros existing. Anyway, git was installed via this plugin because it makes the install and maintaining quite faster and easier. It did not ship with Unraid 6.6 as default, so I chose the package closest to what I found in the Nerd Pack plugin. smartmontools 7.0 was required because of a json output which makes the life a lot easier, and it was tested with no problems.

No seriously, Slackware is a stupid idea to build anything upon these days. It's not a feature, it is just a pure shitty system with no dependency check whatsoever. And sadly I don't know if that top line can be changed other than manually checking the version before installing/replacing/upgrading - which I won't for the sake of a few months or so (I dunno when), before Unraid 6.7 will be released. This is something I really hate with Unraid, Slackware is a very specific and a weird distro (The most Unix like they call it). I started using Slackware back in my early linux days and used it for years, but it is a nightmare when it comes to packages. The good point though, is that it will probably be removed from the script when 6.7 releases, including smartmontools which is also 7.0 in Unraid 6.7. Good to see that git is included by default in 6.7. As for now I have to focus on the stable releases. So when 6.7 is released as stable, this plugin will likely require Unraid 6.7+ This plugin is still in Alpha stage, maybe Beta, even if it's not really written anywhere. Sorry for the harsh feedback, but I think it is rather a waste of time and resources for something temporary as this is. Even if it could be as simple as just make another install file for 6.7, but I really can't be bothered. Again, sorry

1. Less likely to be implemented, will require a bit too much exceptions and will rather break a lot of configs. Why not use BTRFS Raid1 as Cache? I understand different needs, but it is possible to do it native. This plugin is supposed to work as automagically as possible, so I don't think this will be implemented. Not in the current state at least. Sorry about that. 2. Yes, that for sure can be added in a later release, including all other missing variables visible at the Info page.

That's good!

Update 2019.02.10 Commit #48 - BUG: Warning on faulty SQL statement, corrected.

Update 2019.02.09 Commit #46 - MINOR ISSUE: Warnings related to empty variables, will now check before parsing them. Commit #44 - BUG: Looses allocations when pulling out a drive, probably caused because it still existed in the "location" table. Now it gets deleted. Also checks for an older reinserted device and modifies the status in the "disks" table. Commit #43 - FEATURE: Added a checkbox for hiding contents of the "empty" trays.

Update the just released plugin, it should contain some bugfixes according to slot allocations. Else it will not follow the slots, so you have to reassign the drives you move physically. It is intended as people might change configurations, and when they do they also likely know where to drives are located.

This is now fixed in the devel branch 2019.02.08a-dev, will be released at next stable.

Updated post about stable and development build at: For those who want to test out the latest change: Commit #46 - MINOR ISSUE: Warnings related to empty variables, will now check before parsing them. Commit #44 - BUG: Looses allocations when pulling out a drive, probably caused because it still existed in the "location" table. Now it gets deleted. Also checks for an older reinserted device and modifies the status in the "disks" table. Commit #43 - FEATURE: Added a checkbox for hiding contents of the "empty" trays. The development build can be installed and tested, be sure it says version: 2019.02.08a-dev. Do not install if that version is changed even slightly as it might contain broken changes! Else read the post linked above for clear installation routine.

I think I figured it out, but I can't really test it. But I think the problem is that the old disk is still saved in the "location" database while it got tagged 'r' (removed) or 'd' (deleted) in the disks database (for the history sake). The next update should bring an updated function to this including an additional just in case someone adds the older drive back to the array again. I just hope I'm right

This is clearly a bug and is not intended. It should move away (setting a byte in db) the drive which is pulled out, and the new is supposed to end up in the unassigned list. I will eventually look into this, reassigning all drives after pulling out one is clearly unproductive.

I will make it possible to hide the contents of "empty" in the near future, but it will be either all or none of the empty slots - not selective.

As for the current design this is not possible without some ugly hacks which would not be very user-friendly to deal with. Reversing the numbers would just make them reverse from the last tray, and not from the "current" tray. So a multi column or a multi row configuration would not give you intended results. The future of design 2.0 will likely be more configurable in all possible ways, but that's a future project. So for now you can maybe just hide the "tray number" instead, if the numbering is bothering you a bit.

Ah, well, that can probably be added in the nearer future. But it will be a simple config to hide the text and LEDs in that case.

Hiding empty slots won't happen as it will ruin the current design, empty slots are treated different from filled ones. Only way to hide empty slots is to assign them and skip the visible rows. Also the use of "override" can help in some cases, but it will only be shown as the "last row/group" as it is intended for it to be used for other internal drives etc. It is very hard to fulfill all the different cases out there, so limitations will happen. This might be added as part of the grouped design which might happen in the future, but no current task has been done for this yet. Don't get any hopes of for it, as it will require pretty much a large rewrite of the current system. No support for USB drives, it must be connected and recognized via the "SCSI device" in linux as a disk. Typically SAS/SATA drives with internal or external controllers. The main idea of this plugin is to keep track of all the drives in a typical HDD Hot-Swap Tray configs found on servers or 5,25" modules. These are a bit more critical to know the location of when replacing a drive, as the serialnumber might be a bit hard to read when the disk is mounted and connected. Currently no USB support will be added, as I don't see so much usefulness in it. Correct me if I'm wrong and feel free to come with arguments and situations where it would actually make sense.

Update and you'll have it