Delete the relevant files from the flash drive (/boot/config/plugins/dockerMan/images). Clear your browser cache.
Reboot, and hit the docker tab. The icon should now get downloaded and placed in the above folder. Replace the icon with what you want. Clear your browser cache again, reboot and the new icon should now take effect.
Side note that on 6.8.x the icon system has been overhauled and will always take effect if you change the IconURL in the template (template authoring mode in settings - docker)
If they already had it installed, then an update will not change any paths which the user already had. It will however add in any paths which are new
Then they're adding the app back in via previous apps which will reinstall the same template they already had. They would need to add it in from the new apps
Here's the starting point for a command. But, you'd also need to append the current csrf token (/var/local/emhttp/var.ini). But even there, I have no idea if the command would work under 6.8 since you're not logged in either.
My best suggestion would be to place those commands for the download in the go file, prior to the emhttp line (with appropriate waits to ensure the network is running)
https://forums.unraid.net/topic/50490-solved-commands-to-start-and-stop-array/?tab=comments#comment-496967
Implies that privileged is required
These are the capabilities that privileged mode enables. https://linux.die.net/man/7/capabilities You can however cherry pick what you want via some extra parameters, but it's generally easier to simply check the box off.
I've never kept it stopped long enough to see what it does. Not much point in running unRaid if the array is stopped. Better off to shut it down / leave it running but have the drives spin down etc.
If you move a file from
/mnt/disk1/someShare/file.mkv to /mnt/user/someShare/file.mkv, the file will get corrupted. Damn near impossible to do over the network unless you enable disk shares (disabled by default)
A cache-only share, and an array only share are identical are are all located within /mnt/user. IE: shares are shares are shares. On what disk they are located is immaterial to the system. Only when you mix a share and a reference to the disk do you run the risk of corruption if you don't know what you are doing.
You can search for "User Share Copy Bug", but it is not a bug in the system, but a lack of understanding on the users part and as noted very hard to replicate unless you're working directly at the command prompt in which case you should know what you're doing in the first place.
It spins them up so that it can issue a sync command
In the GUI (Settings - Disk Settings) it can be set to between 15 minutes and 9 hours
I use 15 minutes. It's all up to you.
That's fine, no problems. It's still share to share
The easiest way to do this and to have everything in the webUI all in "sync" with the updates etc would be to install the Auto Update plugin.
There you can configure it to automatically do this on a schedule etc. But, if you still want to trigger it via the command prompt, then
/usr/local/emhttp/plugins/ca.update.applications/scripts/updateDocker.php > /dev/null 2>&1
Edit: You will have to configure the plugin though as to which apps to update, etc
Doubt it. To do that after every file is transferred you would have to then read the file from the array to calculate its CRC.
If you don't trust your drives, then you should replace them.
If you don't trust your memory, then you should have ECC
Admittedly I'm confused. Why would you want to start the array via a command in the "go" file instead of simply going to Settings - Disk Settings - and enabling auto start?
6.8 solves the issue of concurrent writes to the cache drive significantly degrading array performance. You should install it to see if the issue goes away.
Any write error would have the effect of disabling the drive in question. But, if there are other underlying problems (eg: bad RAM) that can affect the integrity of the data being written, then the data may or may not be compromised during its transit.
because @kubed_zero has the plugin attempting to download the png and an auxillary script for it every time it installs (and at boot time the network is nowhere near running yet) instead of including those downloads within the ,plg itself
Docker technically starts first, but in the very small delay before the containers start, kvm will get started.
I'd still try safe mode to see if it happens there.
I'll assume that you've rebooted already to no avail.
Try uninstalling SNMP (and reboot) and see if that makes a difference
After that, uninstall NerdPack (and reboot)
If still no difference, restart in safe mode