Squid

Looks like it's been like that forever. Personally though, for any custom cronjobs I add in via a plugin, I check to see if the array is running before the script executes. Don't know if the built-in scripts perform that operation if they require a running array

Add a new script. The script will be simply /usr/local/sbin/powerdown -r Have it run at a custom schedule as this: 0 3 * * *

ENV PORT 443 ENV REDIRPORT 80 ENV MPSPORT 4443 ENV EMAIL mail@host ENV HOST host.ltd ENV SMTP smtp.host.ltd ENV USER smtp@user ENV PASS smtppass! ENV DB netdb ENV MONGODB "mongodb://127.0.0.1:27017/meshcentral" ENV MONGODBCOL "meshcentral" Add a new variable for each of the above (variable and the name is the 2nd item, its value is the 3rd). You'll need to adjust the values accordingly for your setup EXPOSE 25 80 443 4443 27017 27018 Add a new port for each of the above container ports, mapped appropriately to a host port Then you should be able hit the shell for the container and do whatever it does according to the instructions

You want to leverage the inotifywait command https://linux.die.net/man/1/inotifywait

And there will never be. Problem is that when reading the file from /mnt/user/share/..., unRaid will give you the version from the lowest numbered drive (cache, disk1, disk2, etc) The version which you may be expecting, or which may be the latter version could be on the highest numbered drive. Hard to say, and I don't want the plugin to attempt to guess what the user wants to actually do. You can manually clean them up via Krusader (if you map /mnt instead of /mnt/user in the template) by navigating to the appropriate drive(s) and inspect the files accordingly Should be noted that under normal circumstances, duplicated files do not happen under unRaid.

In my case, muscle memory is the primary reason (I'm so used to typing /tmp/GitHub/...) to hit a desktop share. Secondary reason for me at least is that (and I could be wrong here - can't confirm as someone is currently watching a stream) is doesn't UD umount everything at array stop? I don't necessarily want that. But don't change it for me... I'm an edge case

Fix Common Problems' Extended Tests looks for duplicated files. (/mnt/disk1/share/test.txt and /mnt/disk2/share/test.txt)

Wow. ok. Must be my brain frozen then (work sent me to a place where the HIGH is -25.6 F)

/mnt is out because unRaid potentially manages everything in there with the exception of /mnt/disks. Seems to me though (without reading the thread) that it's an edge case scenario, so I'll keep FCP recommending /mnt/disks, especially since if you're manually creating a mount point you should know not to have it interfere in the first place with any mounts created by UD.

Too late to delete my post then, eh? Did that change semi-recently then? I swear it was different before, and had to TV into network to verify

Really? Did not know that It should be noted that if you add a mount point at /mnt then FCP will complain that it shouldn't be there and recommends to use /mnt/disks instead

You gotta love honesty

Yeah. One of my "Issues" with UD and SMB is that the order of the fields is wrong. Before hitting load shares, I always had to go to the username / password fields below the button. Couldn't be bothered to ask you to change the order.

Any long running processes spawned by the actual script do not always wind up getting killed. The plugin does attempt to kill off any child processes of the script, but depending upon what that child does (ie: spawns another child) then it does not always work. Creating another script that has your pkill command in it should however work correctly. EDIT: Looked at the code, and there *might* be a bug in user scripts attempting to kill the childs. Will have to investigate further. How I'm using pkill doesn't jive with the man page on it. Can't remember exactly how I came up with the syntax I'm using.

Probably in retrospect, this all makes perfect sense as user scripts doesn't execute any script through the shell, so any environment variables aren't set. (This is also why you have to give items the full path in some cases vs when executing via a shell you don't have to)

Then there were probably 2 results, and assuming that you're running the latest CA with the new GUI, then you're the first to find a minor oversite - on the main display's it isn't showing that one of those 2 results is "private" (ie: the install from dockerHub). *Alternatively, and I'd have to test is that it only displayed one, but because of the identical naming and repository it displayed the private one. In which case you've found a bug in CA Go to statistics, click on Private Templates and hit the red-x on anything you're not actually using (esp. krusader) and try the whole thing again.

Were you in previous apps by chance? That would install the same template as before. You need to do it from the category list or via a search for krusader

*technically* 2 of the icons do have color - the pin icon changes color accordingly, and the exclamation mark icon changes color from white to orange to red depending upon what extra info I'm trying to convey to the user that isn't in the description.

Don't misunderstand me. My opinion did change after using 6.7. Even forgetting the far far better dashboard, the entire setting's menu is cleaner and nicer to work with. But, feel free to post accordingly. Everyone's opinion counts.

lol. Once 6.7 started down the pipe I also complained about that. Personally, I didn't see how a modern UI is nothing but a retro look to monochrome / lowres monitors. That being said, I'm also the first to admit that I have no idea what a UI should actually look like. Personally, since my primary server is running 6.7 and the backup is running 6.6, I find the 6.6 one looks absolutely terrible compared to the 6.7. (And controlR / cachedirs looks completely out of place) They're only blue because you're not running 6.7 After you upgrade, they'll be white like everything else. Realistically though, the proper place for these kind of complaints is in the 6.7 thread or the feature requests thread. As you've said, I'm simply following the herd / conforming to the design of the UI as a whole.

Did you install it via "Get More Results from docker Hub?" Your screen shot show no icon for it either. Remove the image, then reinstall it by searching for Krusader. The template has both an image and a webUI entry set.

Previous Apps

You just gotta always be right... On next weekends release

All depends upon the size of the browser, the zoom of the browser, the font-size dynamix is set to. Not a real biggie to get it to always stop at a complete line, but the complication arises if you say resize your browser which changes the math. Now, what page are you on? Do you start displaying the apps based upon assuming you were looking at the top of the page, or were you looking at the bottom. All in all, easier to give it a set number (25 on computers, 10 on mobile devices).The 25 though admittedly is a hold-over from the previous GUI where 5 across did fit on my screen (when running 1920), but it's a nice round number that didn't bother me in the least

No clue. There is nothing wrong with the template, and it works on my system. Perhaps there was something else going on with the docker subsystem that was being hidden from view. Either way, glad you've got it going.