Squid

Extended Test does not fix anything. (Because it does not know if the permissions are what you actually want) Docker Safe New Permissions fixes the permissions on everything but the docker appdata share, and any additional shares you set.

Add it again, but name it differently and assign it different host ports.

Did you add another path mapping to the docker container? (ie: /MyNAS mapped to /mnt/disks/whatever) With an access mode of Read/Write:Slave?

Stop the service, then delete the files, then move, then recreate

Added in (actually, @Djoss did) checks for if a user changes the TemplateURL within a docker's template (if the author specified one). This test also includes an automated fix for these problems if found.

Admittedly, I'm not in the "right frame of mind" at the moment, but I'm sure that sure that it does that automatically without any option to disable. Assuming of course that you have set up notifications properly in unRaid.

To be quite honest, I didn't look into your situation per se (as I knew the code was solid). Rather I thought about worst case scenarios, and adjusted the plugin to account for that. Believe me that I got very annoyed at just how long it takes for unRaid to boot if it has no internet / can't resolve DNS. Luckily my wife wasn't around this morning to have to deal with that

@RobJ's ideas Bait Files are more secure, but more prone to inadvertent trips. Bait Shares are more convenient and less prone to inadvertent trips Your #1 attack vector by far is via SMB. NFS, a while ago I had asked for help on how to make a NFS share read-only in unRaid's implementation, but nobody could / would answer the question (short of modifying the permissions on the files which I don't want to do) If the automatic backups delete any files in the destination share that doesn't exist on the source, and those files are bait files, then yes. An attack is an attack. Bait Files in all directories and Bait Shares. Unless you're in 100% control of the files on the server though, Bait Files in the root only with Bait Shares. If you don't have 100% control (other users regularly accessing / deleting their own files from folders, etc, Bait Shares only are your best bet). My self, I use 20 Bait Shares (containing ~1,000,000 bait files) and bait files within root of shares only. If a malware author was going to try and work around this implementation, it would be nothing for them to ignore the default naming of the Bait Files. But, you can change those how you wish to avoid that problem. But, you have the option to create and use your own Bait Files. Bait Shares. Anything is possible, but the file names are all randomized, and are named in a variety of different ways that IMHO simulate the naming of files that would be present in somebody's work computer. Would any of the authors go through the trouble of trying to get around this plugin? Doubt it, but you never know. On an attack, there is an attack log that does specify which particular file was attacked. Best that I can say is that of the 2 users that I know of in this forum that have been hit with a ransomware attack (wannacry), neither one of them were running this plugin. I wish they were.

Changed the behaviour of what happens if a failure to download the application feed happens. Previously, CA would automatically revert to Legacy Mode. This worked fine in the case of the server handling the application feed was down and GitHub was still up. But in the case of both servers being down, or a failure to resolve DNS addresses, then the cumulative timeouts for downloading each author's repository adds up to ~1.5 hours during which unRaid's webUI may be completely unavailable. Now, if the application feed times out or is malformed, an error is tossed onto the display suggesting to fix any DNS issues that you may have and giving you the option to switch to Legacy Mode (Additionally, in the case of a malformed feed, the feed itself is displayed as an error). This error should only take about 3 minutes tops to appear in this case. Side Note: unRaid sucks at starting / stopping if it doesn't have a valid DNS servers. (half hour on my system to do either )

Your guess is right

https://forums.lime-technology.com/index.php?/topic/39731-[Devs]-Community-Applications-List Tapatalk mangled the url but the thread has the appfeed url in it Sent from my LG-D852 using Tapatalk

Mobile right now at a bbq so not for a bit. Sorry man... Sent from my LG-D852 using Tapatalk

Nothings changed in that respect. I suspect a dns issue. Uninstalling and reinstalling would provide the info I need. But it might take a couple of minutes for the attempted download of the appfeed and the subsequent automatic reversion to legacy mode to time out before the webui responds again Sent from my LG-D852 using Tapatalk

Delete the orphan

Remove the container from within the docker tab, then re-add it either via the template drop down (user-template: my-plex), or via Apps Tab, previous apps. Either way, after the download is done everything will be back to normal.

If you google all this, there are ways to stay on your current network (and access the internet without going through the VPN) and remain connected to the VPN

Changed New/Updated section to include sort by date. Automatically switches to Date sorting (Descending) when entering the section Descriptions in New/Updated section include the Date Updated Sorting Reverts to Name (Ascending) when leaving the section Fixed Selected Category not resetting after a search performed Fixed regression error where if updating CA from within CA, the page wouldn't reload after update installed * * TLDR: Long story. Even a summary is a long story.

https://github.com/Squidly271/ca.documentation/blob/master/caDocumentation.html

Updating the manual is on the todo list, but I keep putting it off as documentation is by far the least enjoyable part of any plugin... But I will get around to it....

Removed it on the last update My reasoning in this link: https://forums.lime-technology.com/topic/38582-plug-in-community-applications/?page=66#comment-570490 You're far, far better off installing cAdvisor through CA. (lol And with all the GUI changes to CA in the last few months, most of the vids out there are way out of date - including the OP)

Just type mcelog at the command prompt. If you're running the Fix Common Problems plugin, then it will log the output of that command during the next scan if it finds any mce errors

Ok. You caught me in an inconsistency. I left New / Updated which uses the date, but removed the sort by date as authors are terrible at updating the flag. I guess that I've got two choices; Remove the New/Updated Category Reinstate sort by date on New/Updated Probably best if I pick the second option () When switching to that section, I'll have it automatically switch to Date Sorted (Descending) which will pop the new apps up to the top of the lists. But, with the prevalence of auto-updating apps, the docker app guys don't ever change the date updated flag to match the auto update version every Friday at 2300GMT (and I do understand why), I'm going to leave the description / pop up the same as I changed it to on the last release, namely that it won't display the date updated as text, because it's very confusing. (And at this moment, I can't cherry pick what to display within the various categories.)

You will need to switch to Bridge mode. There may also be other ports beyond 32400 that you have to map

At the command prompt, type in mcelog Or run a scan again, and FCP will bitch and complain again, but this time the log from mce will get logged into the syslog so that it'll appear in diagnostics

I'll take this as a good sign and remove the mod comment within CA on the plugin...