ich777

SLI in a VM is only possible with hacks IIRC but I have not tried if this is even working. I used it back in the day on a non supported Motherboard with Clover and a custom dsdt file where I injected a few lines so that the driver thinks that my Motherboard is SLI compatible. I've written a tutorial on win-raid.com back in the day but the Forums there are not available anymore...

May I ask if you have the same entries in your syslog or do you have different errors?

It should be enough to create a Device entry like this: But I have to say I'm not too sure about that since I don't know how they monitor the device. May I ask for what monitoring is needed? Have you tried yet to issue this command form a Unraid Terminal: echo 2 >/proc/sys/kernel/perf_event_paranoid and then add the cap: --cap-add=CAP_PERFMON to the container? If that doesn't work then try to set it to: echo 1 >/proc/sys/kernel/perf_event_paranoid (but I don't recommend to set it to such a low value) You have to restart the container after you change the values on Unraid itself.

I have had theses issue all the time even with the old NUT package (v2.7.4) as well as with 2.8.0 that @SimonF ships with the new package and with my package too. This is for example from a few days ago (with package v2.7.4): Jul 23 13:34:29 Server usbhid-ups[4779]: nut_libusb_get_interrupt: Input/Output Error. and this is with the new one: Jul 27 00:16:11 Server usbhid-ups[12858]: nut_libusb_get_interrupt: Input/Output Error I have tried many different cables and it's always the same but I'm not to worried about that since everything is working. Or do you talk about another error? If yes, please post screenshots or the messages. Yes, it will also tell you which version I used when you do a: /etc/rc.d/rc.nut reload You should see something like this: Network UPS Tools upsd 2.8.0-signed Network UPS Tools upsmon 2.8.0-signed No worries, of course not, just answering questions and explaining how I do things.

Not really. It seems to me that the other package is compiled differently how it should be. Please look again at my package and go into /usr/libexec/nut (this is the correct path) and you will see all the drivers (at least that's how NUT calls them). I don't know how the other package is compiled but when you use the flag: --with-drvpath=/usr/libexec/nut the drivers will be installed there. This is because I don't include the doc and man folder (no man available on Unraid). May I ask if something is not working? I use it since yesterday and everything is working fine over here.

I think your issue was completely different that your system became unresponsive and that transcoding stopped working or wasn't working at all, even on native clients. IIRC you also had some entries in your syslog from the driver where it stopped working correct? If yes then that's not the same issue. Have you yet tried native clients (Android, iOS, Samsung)?

Please tell me this is a joke... A few users lately reported this issue to me and I also think this was a common issue a few years ago right? Will Plex fix this issue or not? My issue with that is that I currently have to compile 4 or some times even 5 drivers for each new Unraid version and I think you can see that this is time consuming, even if it's an automated process: Production Branch New Feature Branch Beta Branch Open Source Branch Legacy Branch Sure thing, I have to recompile the old driver but I even don't know up to which Kernel version the 525 driver is supported but I also don't think that even that is true that only the v525.x drivers are working since the legacy driver v470.x also doesn't work (reported by a user). Also, I read one of the comments from a Plex Team Member in the linked post: Here are the drivers: https://www.nvidia.com/download/driverResults.aspx/204837/en-us/ Did the Plex Team ever say why it is not working with the newer drivers? I really don't understand why it is a driver thing when it is working on native clients but not on the web client, this has to be an issue with their web client from my perspective and not with the driver, hope that also makes sense to you. @alturismo have you heard of that yet? Sure, I have to compile the v525.x drivers for each Unraid version but I really don't understand why this should be a driver thing that the web client on a maybe far far away machine is not working?

I hope you don't get me wrong here but I've not gone through the entire rc.nut file which ships with the plugin but even in the rc.nut that I ship with the 2.8.0 release everything should work fine. Hopefully you get my point here, I really don't want to mess with permissions which developers maybe intended to be differently. Maybe I have some time to go through the whole rc.nut file and why it is even shipped with the plugin instead of using the one which ships with NUT itself. Anyways thank you for your PRs and all that you contribute here.

Because that's the correct permissions and how the developers of NUT have intended them to be, maybe not for chown -R 218:218 but for everything else these are the correct permissions and how the are compiled. The files which are in /etc/nut should have 644 and have them by default, also why should a config file be executable? That means that all the files in this directory are changed to read only recursively. Usually on a Linux filesystem you don't set all permissions to 770 to ensure that everything is working, at least I'm strongly against 770. For Slackware I would recommend the permissions 755 and not 770, there is even an option when creating a Slackware package to set all permissions to 755 and even there it's not recommended doing that. Hope that makes sense to you.

I‘ve now gone through the PR, please don‘t change the permissions to 770 they should be set to 755 Also I wouldn‘t recommend changing the permissions at all.

This was done on purpose and for two reasons. When I added this to the plugin I was working on small screens remote and I was constantly hovering over it and it kind of bugged me that this always pops up and the second thing was that it made no sense to me to show a tooltip that shows exactly the same. The whole point for the footer style Small was to make it minimalistic, maybe I should have call it Minimal instead of Small.

This seems about right. If you changed nothing and just let the container just start it is somewhere on the list because this dedicated server will actually stop when there is something wrong with the network configuration and the dedicated server won't even start. My server shows always up on top of the list in game. I guess maybe there is something else preventing you from finding the server on the server list.

Everything seems correct, where do you have a issue? Is this an old instance from the container or did you just install it? How much RAM and CPU does the container use? You can ignore for example the GPU and shader errors because a dedicated server has no GPU or doesn't need any shaders to work. I just tried it on my server, I've attached the log from a completely fresh install which is working perfectly fine, at the bottom you see this message: #DSL Dedicated server loaded. Here is my log: sotf.log I can also connect to it just fine: So it looks to me that you server is also just working fine and you should be able to connect to it.

Please upgrade your Unraid version to something more recent. I would recommend that you upgrade to 6.12.3 or at least 6.11.5 This should resolve your issue. Staying on a lower version is pretty risky and I even can't give proper support because you even can't get that version anymore.

Please post your Diagnostics.

Thank you for the report, I will deprecate this Docker container because they are constantly changing the download link from the dedicated server executable and I'm really tiered of changing it around every half year. Sorry for not having better news for you.

Then the ports to the outside world are definitely closed or something else is going on. Did you also change the ports in the template or read the description from the ports at the template? If you change the ports in the config you also have to delete the existing ports in the template and create new ones with the Host and container port set to the same as in your config, of course in this case with UDP.

No. I think there is something like nut-cgi out there but this is not part of the plugin. Why would you want a WebUI when you already have one in the plugin?

You talked to MariaDB? If you have searched the issue it is pretty clear that something is still wrong with the database. Have you created tables in your MariaDB database? It seems that the Database is not empty and therefore it thinks that every table which is needed is in place but it is not and basically it tells you that some tables in the Database are not found. I would strongly recommend to start over, maybe with a fresh Database...? Do you have anything else than Zabbix in the MariaDB container? It would be also helpful to know your config and what you did to create the Database, grant access and so on.

If that doesn't work I would recommend that you go to the Support thread for the MariaDB container that you are using and solve this issue first. As @alturismo has already proven, Zabbix is working. Now it's up to you to create a database/user, grant access and connect Zabbix to the Database.

@blub3k, @CoolHam, @moarSmokes, @bailey & @Oasistem: The issue with the nvidia-smi exporter should be resolved, just make sure that you are on plugin version 2023.07.23 and everything should work again.

@thecode I don't have that issue over here with Manual: If I change that back to Auto it looks like that: You might have to wait a bit until the Footer/Plugin page updates the values depending what refresh interval you've set.

This is not a good idea since of you are using the database for something else it will definitely mess up your tables and cause issues. My recommendation is that you set up a dedicated database in your MariaDB container and a dedicated user for that database or you go the Docker way and set up a whole new dedicated MariaDB container for you Zabbix instance. I would recommend that you look up how to set up a database and user in MariaDB and grant access to that database to that user and the correct IP address where Zabbix is running on. For example: Click (please read the whole article since it mentions also how to grant access to non localhost <- you need this) This seems like a MariaDB issue (or at least some kind of problem) and not a Zabbix container issue if you can ping the MariaDB container from the Zabbix container.

Yes, only pro, please look at this post:

As @alturismo said, that's most certainly a Plex issue. From what I see in your Diagnostics the driver was load fine. BTW may I ask if you use the card in another container too since it looks like it was detected fine in another container or even the Plex container (can't tell for sure in the logs since there are no container names listed). Through what client are you trying to transcode? If it's the web client please try a native client like for iOS, Android, Samsung,... since the web client most of the times is not working when you switch over to transcoding.