Last night I rebooted my system to get the latest updates. When it tried to come back up, I got the error in the title. As luck would have it, someone else had the same error last year. Note only was it the same error, but it was for the same use case: running a dell poweredge with nvidia drivers. I was able to fall back to an older kernel for now, but I would appreciate some advice from one of the experts as to what my options are. I am a data scientist that uses my nvidia card for machine learning and other tasks, so I would rather not take the card out of the equation. It was a bit of a pain to get the nvidia drivers working in the first place - I am hoping to not have to mess around with them again. I am just speculating as I don't understand much about this area of these machines.
Posts by muchgooder
-
-
And it turns out it's my fault. I had installed Nvidia drivers for a transcoding GPU and that seems to have broken the kernel upgrade leaving it unconfigured. sudo apt remove nvidia-* and all is well. (Except for transcoding but that's a different problem)
You just saved me a boatload of time trying to figure this out. I also have a poweredge and I had also installed the nvidia drivers so that I could perform machine learning using my video card. Were you able to come up with a long-term solution?
-
In the video that I saw, he didn't mention anything about uids for docker. I don't know what he may or may not have said in the past, but his current video is very well done and didn't go against anything said in the docs (I had them open as I followed the video).
Yup, I will definitely need to take care of the UIDs and paths.
Ok, so it would make the most sense for me to use the UI if I want to see the service? That is what I was looking for.
-
As always, thank you to everyone that participates in this forum to help out the weekend warriors.
I created a fresh install of OMV7 on a new box and that went perfectly. I then went through DBTech's tutorial for setting up docker - that also went off without a problem and I was able to install Docuwiki as a test.
At this point I am trying to port over my configuration from my OMV5 server. On that server, I had all of the individual containers set up in their own folders under /AppData. They all have docker-compose files. I want to have the same setup on the new server. The question I have is around best practices and ease of use. I thought that I could simply copy over the entire directory, edit the paths, and then issue docker-compose up to launch the container. This works, but I noticed that taking this route results in the container NOT being registered under "Services". I do see the container as being "up", but I can't stop or restart it from the UI.I did some research and I see that there are a couple of "Import" options, but after reading the docs, I am not sure that this is what I want.
So the simple question: Is there a way to quickly and easily import all of the folders from OMV into the new server and have them show up in the UI with the ability to start/stop/restart? Or should I just copy over the config and paste the contents of docker-compose into the UI (I am assuming that it manages it here?).
-
Yes, that looks like what is done by OMV when first installed.
Ok, thanks for that! So my best course of action is to get a bigger SSD? Or under normal operation does it only take as much as it can get for system activities while leaving some free space left?
-
I noticed today that my OS drive ran out of space, so I took a deeper dive into that drive. Does OMV partition the drive like this?
-
I don't know what the original issue was, but the secondary issue (drive full) was solved using this guide: [How-To] Fix full OS filesystem - GUI login loop
-
During this last week, I swapped out two DATA drives (not the OS drive). I have a rack-mounted server so the process of removing the drives was easy (e.g. I did not have to open up the server to pull out the drives). I used a disk cloner to make the new drives, and the server was back to normal a few days ago.
I was hoping to make some changes to the shared folders this morning but I could not log in to the admin panel (it would appear to accept my login, but I would be brought right back to the login). So I rebooted and saw these errors. It has been 20+ minutes and I have not seen the end of them.
EDIT 2: I disabled fast boot and re-seated the drives that I add this week, and the box came right up with no errors. Weird. However, I am still not able to log in. Upon further inspection, it looks like my OS drive is full. Still working.
EDIT 1 : It looks like it is still working on startup activities almost an hour after I rebooted it. Many errors involving the network service.
-
I guess that's what comes with OMV5 that is likely not updatable anymore. You could try downloading the current version and try installing it by hand.
I am actually all set with this one. Yesterday afternoon I posted the steps that I used to resize the disk - I used cfdisk instead of growpart. Thanks again for your help with this.
-
That is correct. Check the version of the package it came out of.
.29-1, which it says is the newest version.
-
The versions of growpart I am using come from the cloud-utils package, versions 0.33-1 and
0.32-22-g45fe84a5-0ubuntu1.
I have used these on 8, 12, and 14TB disks with no problems. What version are you using?For some reason it isn't returning the version. Maybe because I installed it as part of cloud guest utils?
-
This is completely wrong. Where did you get that from?
If you google that, you will find numerous posts about it. It may have been specific to certain releases - here is one: https://www.suse.com/support/k…yond%202TB%20successfully.
-
I do have some good news to report, and hopefully this helps someone in the future. I was able to resize the drive. Thanks again to everyone that participated - I can't tell you how much I appreciate the support of the community (and I promise to upgrade soon).
Note that I did not unmount the drive, although i agree that it would have been safer to do so (this drive contains movies and shows) -
Have you determined that an 18TB drive is compatible with the system BIOS of the machine?
Yeah, I did look up the motherboard specs before buying it. Incidentally, it looks like growpart might only work on drives that are 2TB or less (still looking to confirm that).
So I did have a WTF moment a minute ago. I rebooted it again for the hell of it, and it came back up with the right docker path. I have absolutely no idea why. It came up and everything was as it should have been.
-
You can create a Linux Live CD or Live USB and boot with it. The Ubuntu or Mint distros include Gparted, as do most others. I find these things painfully slow to boot up but once booted they are workable. Have patience.
Great, thanks much - I still have some of those laying around from 15 years ago.
-
Again, it is better to post the actual text, not a picture.
Is there more than one partition on that disk?
You may want to attempt this with gparted in a machine that will work for it.
No, I do not have partitions on any of my drives.
Is there a program that will allow me to do this in windows? -
Based on your output, the repo problem should be fixed now. Are you still seeing the other issue?
Which issue?
I suspect that I will have a docker issue until the new drive comes in. But I am back to trying to re-size the cloned drive that I dropped into the server. -
Most devs I know want bleeding edge everything lol.
What is the output of: sudo omv-changebackports NO
Trust me, I am that guy. But I became a data scientist last year and all of my free time has been spent ramping up (and keeping up) in that area. I kind of left this alone because my home runs on docker apps and everyone was happy.
Anyway, that command ran with no issues.
-
But you should be upgrading from 5 to 6 to 7.
I will help you fix the repo to upgrade not so you can stay on 5. I struggle to remember how omv-extras 5 works even though I wrote the code.
What is the output of: sudo omv-aptclean
Thank you so much! Just to be clear, i don't know if that repo is actually causing an issue. I just know that when go to do a variety of operations that I am greated with ugly errors that seem to reference the repo. I've been waiting for a sunny day to migrate... it just hasn't happened yet. I am also a developer and usually like to adhere to the "if it ain't broke" philosophy, but I know the downside of that with OMV is that people will (understandably) not want to help with older versions.
Here is the output. I capture the bottom of the output as everything was fine until then. Again, I have no idea if this is causing any other issues. I know that my docker drive is bad and that will be replaced with a new drive tomorrow. -
Sorry, I can't help you with upgrade 5->6 problems. It may be very difficult or impossible to do this as OMV5 went end of life quite some time ago and its repos may be shuttered.
I think you misread my post. I am not trying to upgrade from 5 to 6. When I perform various operations, I am presented with the big ugly error that I referenced earlier, and that error came up in posts where people were trying to upgrade from 5 to 6 (which I am not). It seems to be referencing OMV Extras.
