kennymc.c

The Pushover API offers the option to set a sound parameter in which either predefined or custom sounds can be played when a notification is sent (https://pushover.net/api#sounds). It would be nice if this parameter could also be set in the Unraid GUI. Additionally, if a sound is selected, the priority of the test message should be set to 0 instead of -1, otherwise Pushover will not play any sound. I already tried this by hardcoding -F "sound=gamelan" \ after line 26 in \flash\config\plugins\dynamix\notifications\agents\Pushover.sh.

Through a Github issue in the Telegraf repo I became aware of an alternative Telegraf Docker image (golift/telegraf), which is based on the latest tag but additionally contains smartctl (smartmontools), ipmitool, nvme-cli, sensors (lm-sensors), mtr (mtr-tiny) and sudo. For everything to work use_sudo = true must be set on all required inputs that support it. For monitoring Docker, you can add --user telegraf:$(stat -c '%g' /var/run/docker.sock) to the Extra parameter field in Unraid, since it works without sudo.

I think ailliano is looking for some packages which are not to be found under this link. iPerf3 for example. You can find more packages here: https://slackware.pkgs.org/current/slackware-x86_64/ I assume that Slackware Current is the same as 15.0 since this should the latest release. But I haven't tried it out. Maybe someone could confirm this. If not I will probably backup all txz files from the nerd/dev pack plugin folder, uninstall both plugins and place the files in /boot/extra after the upgrade and install them manually.

How do I find out which packages were installed via the plugin and which come from Unraid itself? Filtering by installed shows both variants. I guess the packages where an update is available but which are not updated after clicking on them are the ones from Unraid? Unfortunately the page linked a few pages ago (https://slackware.pkgs.org/15.0/slackware-x86_64/) doesn't seem to have all the packages included in the Nerd and Dev Pack. When I filter for Slackware Current I find more packages like the one for iPerf3. Current should also be 15 and compatible with Unraid? Or is there a difference?

I also noticed this problem with the latest preclear plugin update from Monday (2022.07.11). So it is not possible to update the plugin and reboot. System log looks like this when i try to run a script in the background: Jul 13 08:53:25 smartserver emhttpd: cmd: /usr/local/emhttp/plugins/user.scripts/backgroundScript.sh /tmp/user.scripts/tmpScripts/backup_usb/script Jul 13 08:53:25 smartserver atd[11341]: PAM unable to dlopen(/lib64/security/pam_unix.so): /lib64/libc.so.6: version `GLIBC_2.34' not found (required by /lib64/libresolv.so.2) Jul 13 08:53:25 smartserver atd[11341]: PAM adding faulty module: /lib64/security/pam_unix.so Jul 13 08:53:25 smartserver atd[11341]: Module is unknown

Great news:

Is there a logic which files are downloaded directly with one click? For many other files that are not in the editor.cfg, a preview tool opens, but of course it cannot open everything (Unsupported Preview). Is there a way to set which files are opened and which are downloaded? With binary files it makes no sense to open them with the editor. Also the following features would be very helpful: - Upload via Drag & Drop - Download of all marked files by right click or a download button in the bottom bar. Currently only the first marked file is downloaded.

But i think i have to create my own docker image for that, right?

They are executed by the program inside the container, not on the host itself. Without executable permissions on the host system they can't be run inside the container. I also have several scripts on my storage that I don't run on Unraid or in a container, but on the machines connected to the share. I don't understand why you generally shouldn't have anything executable on your storage. This is a security risk in certain environments but as long as the read and write permissions are set correctly i personally don't see a big problem with it. It's not about executing scripts but about setting the execute permission with the plugin.

One reason that comes to my mind: Custom scripts needed inside a Docker container.

The permission tool seems to be limited to changing read and write permissions. Is there a reason why I can't change executable permissions? Can the column order be changed to have the view column on the left? Or is this something that would need to be changed in Unraid itself?

Thanks for the info. I thought so but another user had referred to this thread in a feature request.

The developer has just confirmed the log highlighting is not controlled by his plugin and it is build into Unraid. This is also in line with the statement of another developer who called the new docker/vm log view a design decision. So I think the only way to get the color highlighting back for the docker and vm logs is upvoting this feature request.

Any chance that another developer could continue the support? The functionality is built into Unraid for serval years and i miss this function with 6.10. At least I am not aware that I have installed a plugin for this manually before. I am wondering why the system log in the tools tab still has colored log messages with 6.10. Would it not be possible to continue to display the docker and vm logs in this way, or at least offer this as an option?

Good to know but then why does this still work with the system log in the tools tab?

If anyone misses the colors in the logs I created a feature request for it. Unfortunately it was a design decision

Since the 6.10.0 update, logs are only displayed in black and white via the logterminal pop-up window. Even though this might improve the readability at first glance, the different coloring of info, warn and error messages have unfortunately also disappeared with this change. This made it very easy to see if something is wrong or to filter out relevant info. For me and probably many others, this was a crucial feature that I now miss very much. Now all I see is a pile of text that becomes even more unreadable on longer lines due to the larger font with line breaks. If you go to Tools/System log you can see what it looked like before.

I stopped the container, deleted the folder in /tmp and created a new one via terminal which had a 777 permission. After a server reboot the same folder had a 755 permission. I then deleted the docker container including the image, recreated it but still got permission errors. When I changed the permissions back to 777, everything went back to normal. The affected files and directories are also owned by root inside the container. When opening the console of a container a am always root and can access these files or folders. I assume that the processes inside some containers maybe are not always running as root and they can't access these folder or files. So it depends on the use case. Before the update is was somehow possible for non-root users inside a container to access all files from mounted volumes even if they were r/w/x by root only. At least this is what I can think of as a possible explanation why I am having problems now.

Has 6.10 changed the ability of docker containers to access files owned by root on the host? I noticed some containers were not able to access root-owned 700 or 755 files or directories after the update, like ssh keys or mounted volumes inside /tmp. When I change the permission to 755, respectively 777 for directories to write in, it worked again. Since you can define a docker volume to be read/write or read-only inside the container I am wondering why I now have to set additional permissions which were not needed before. After a restart my sub-directory created in /tmp for a container ramdisk got set back to 711 which again caused error messages. Is there a way to prevent this or was this done intentionally for security reasons?

I can confirm this for the docker logs and was also wondering if this is a bug or a feature. The system log still has colored messages when opening it via tools but not when you open it via the top bar.

I noticed that when you upload a file with the same name that is already in the current directory, it will replace it without any warning. I think this is very dangerous. I also can't see the switch for advanced and basic mode in the top bar that is mentioned in the pdf from the start post. F2 also does nothing.

Has anyone tested if the Docker json/logs files ramdisk script still works with 6.10 stable?

I am currently trying to track all writes to my cache SSD and move what I can to a RAM disk. I have already mounted the /tmp directory in the container to /tmp and thus have seen fewer writes. However, when duplicati backups are running I still see several GB added that I can't explain. The uploaded archive files are much larger and also the databases in /config are combined smaller than the regular increase from the backups, so it must be other files. Does Duplicati write to directories other than /tmp and /config during backups?

Erstmal danke für die Tipps. Ich habe gerade das RAM-Drive für die Docker Logs angelegt. Soweit funktioniert auch alles und die RAM-Disk wurde korrekt angelegt. Mir ist allerdings aufgefallen, dass die Docker Container mit Healthcheck in der erweiterten Ansicht nur noch (health: starting) statt wie zuvor (healthy) anzeigen. Lässt sich das irgendwie anpassen? Edit: Hat sich erledig. Nach einem Reload wird jetzt auch der richtige Status angezeigt.