Thanks Andy! I have no idea! I got this server in 2013 and have downloaded updates regularly, if less consistently. I might have overlooked something. I know I haven't been paying enough attention to this server, which is something I decided to change, but I don't think I've had issues upgrading since 2014. That's... embarrassing given that I work in security. :-) (I don't sell my services as a Linux server expert. But still, it's bad. I should have known better.) I think so, but I didn't write that down, so I can't be 100% sure... ii linux-image-2.6-686-bigmem 2.6.32+29 i386 Linux 2.6 for PCs with 4GB+ RAM (meta-package) rc linux-image-2.6.32-3-686-bigmem 2.6.32-9 i386 Linux 2.6.32 for PCs with 4GB+ RAM ii linux-image-2.6.32-5-686-bigmem 2.6.32-48squeeze19 i386 Linux 2.6.32 for PCs with 4GB+ RAM ii linux-image-686-bigmem 2.6.32+29 i386 Linux for PCs with 4GB+ RAM (meta-package) total 27652 drwxr-xr-x 3 root root 4096 Feb 3 07:53 . drwxr-xr-x 22 root adm 4096 Jan 16 14:09 .. -rw-r--r-- 1 root root 1328749 Feb 5 2016 System.map-2.6.32-5-686-bigmem -rw-r--r-- 1 root root 111663 Feb 5 2016 config-2.6.32-5-686-bigmem drwxr-xr-x 2 root root 4096 Sep 26 2016 grub -rw-r--r-- 1 root root 12217326 Feb 3 07:53 initrd.img-2.6.32-5-686-bigmem -rw-r--r-- 1 root root 12217371 Feb 3 07:51 initrd.img-2.6.32-5-686-bigmem.dpkg-bak -rw-r--r-- 1 root root 2373728 Feb 5 2016 vmlinuz-2.6.32-5-686-bigmem Searching for GRUB installation directory ... found: /boot/grub Searching for default file ... found: /boot/grub/default Testing for an existing GRUB menu.lst file ... found: /boot/grub/menu.lst Searching for splash image ... none found, skipping ... Found kernel: /boot/vmlinuz-2.6.32-5-686-bigmem Updating /boot/grub/menu.lst ... done It's been rebooted several times, most recently on 31 October. Never any issues. Martijn

Hi Andy, Oh, it's pretty clear that I should just set up a new Bitfolk server and copy things there, then nuke this one. (Or donate it to a museum of weird virtual servers...) I'm fairly sure I've done at least some of the upgrades, but I haven't kept a log of it somewhere. The likely explanation for this mess is that I have not had the time to pay attention to this server for some time, but have had at times needed to do something specific. Things like running Let's Encrypt. I then found a quick fix, which solved the problem, and I know enough about Linux servers to understand what I was doing, but never took the bigger picture into account. 10.2 Done! Martijn

Hi Martijn, On Wed, Feb 12, 2020 at 06:28:51PM +0200, Martijn Grooten wrote: Right, so it is at least partially upgraded to Debian 10. Just picking on the first few packages I see… […] This one is from Debian 6 but maybe is just because you've constantly upgraded from there. The above set from Debian 9… The above are from Debian 9… And that one is from Debian 10, so again suggests that Debian 10 was partially installed. It's not as bad as I feared though; you definitely do have some packages from Debian 10. Some really crucial ones are still from Debian 9 though. e.g. dpkg and apt. I think Kamal was probably correct in suggesting that you stopped getting kernel updates when the kernel package name changed and have since been stuck on an older kernel. The need for a more modern kernel has probably caused the upgrade from Debian 9 to Debian 10 to fail so it's only partially gone through. To be honest as you ideally also want to go to 64-bit quite soon, I think I would recommend asking for a new account in which you can install a clean Debian 10 and port all your things over across a period of up to 2 weeks for free: https://tools.bitfolk.com/wiki/Migrating_to_a_new_VPS If you didn't want to do that but instead wanted to try to fix this in place, I suppose what I would do is: 1. Install the kernel from Debian 10 by doing "sudo apt install linux-image-686-pae" (this is a virtual package which at the moment should also pull in the real kernel package "linux-image-4.19.0-8-686-pae") 2. Check with update-grub that it finds it and that it's first in the boot order 3. Reboot into it. 4. Complete your upgrade by doing "sudo apt dist-upgrade" The thing is that quite a lot of packages are going to be upgraded and there's no way to stop that happening because it will also happen when you do the "sudo apt --fix-broken install" that it might want you to do first. It will probably go okay but is likely to be a lengthy process and some software might need reconfiguring. So I wouldn't try it unless I had some time to spare to do it all in one go. Maybe it ends up being simple, but on the other hand it might also fight you. It knows already that the package database is inconsistent, and the --fix-broken will try to fix that by installing the dependencies it thinks it needs, but possibly some of those will need a newer kernel so won't install, and you get stuck. It could involve going to: https://packages.debian.org/buster/linux-image-4.19.0-8-686-pae then following the download links to: http://ftp.uk.debian.org/debian/pool/main/l/linux-signed-i386/linux-image-4… downloading that and force installing it with dpkg -i. That then may fail because of missing dependencies, so you'd have to download the dependencies too, and add them to the dpkg -i line. Could be quite hairy. If you do complete the above steps you'd then finally want to reboot again to make sure everything starts cleanly. I would also recommend editing /etc/apt/sources.list and replacing every instance of "stable" with "buster". Otherwise, when Debian 11 ("bullseye") is released the very next time you try to install something it will try to install the bullseye version. It is better to plan the time when you will do a major release upgrade. If you'd like to go with the new account route then just drop support(a)bitfolk.com an email with the information requested in the article. Cheers, Andy -- https://bitfolk.com/ -- No-nonsense VPS hosting