Announcement

Collapse
No announcement yet.

Systemd Adds Feature To Fallback Automatically To Older Kernels On Failure

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #21
    Originally posted by mzs.112000 View Post
    This would have saved me from a time on Arch when I had to boot up using a USB stick, then chroot in to the actual install, then I had to re-install the kernel....
    Of course, then again, maybe it wouldn't have saved me, because I think all the older kernels were deleted so I only had the newest one installed....
    Yeah, this works only if you have more than one kernel installed, so it can fallback to the old one

    Comment


    • #22
      Originally posted by red23 View Post
      But remember systemd sucks because its this all on one thing that goes against so many unix principles. I actually never got this, never had issues with upstart nor systemd and my fist look at a systemd config file was "yeah that looks simple but yet is powerful whats not to like".
      Essentially that post just kicked a skeleton of dead horse.
      So called Unix principles is dead long time ago killed by X server among the first. Linux is not UNIX as a reminder as quoted below.

      Originally posted by Neil Brown
      "Linux" is just a kernel, so stuff in user-space certainly isn't "Linux".
      It also certainly isn't "Unix" as that is fairly dead. It met needs in the 60s and 70s and even to some extent the 80s quite well. But the needs we have today are very different.

      Comment


      • #23
        Originally posted by red23 View Post
        But remember systemd sucks because its this all on one thing that goes against so many unix principles. I actually never got this, never had issues with upstart nor systemd and my fist look at a systemd config file was "yeah that looks simple but yet is powerful whats not to like".
        Actually, it follows unix principles more.

        Comment


        • #24
          Originally posted by rolfen View Post
          This is a sh*t feature. I am running kernel version 4.4 and at some point systemd decides that 4.3 is better for me, and I am left wondering why some driver or software has stopped working, because I don't know about the automatic kernel downgrade.
          This is meant to be used on servers (where the sysadmin will look at logs anyway) and computers where the user isn't aware of the kernel version and just wants a stable system (i.e. most PCs).

          Exactly how unstable is the kernel expected to be, to come up with such mechanisms?
          Hang on boot (so the user or a hardware watchdog manually power cycles after a wait time) or panic (and reboots on its own). Basically when it boots and does not reach login prompt a preset amount of times. It's not something subtle or trying to outsmart you.

          There is already an option to run an older kernel version at boot. It works well.
          Yeah the GRUB fallback.

          Systemd project has their own bootloader called systemd-boot (it was Gummiboot before), this feature is implemented with that boot loader and UEFI.

          Comment


          • #25
            Originally posted by caligula View Post
            Amazing that Arch still has such a naive package manager. Compared to Debian, Arch still pales in comparison. When I used Arch more, I remember one of the most annoying features was that you could basically do full system update even without an immediate reboot, but it would delete all kernel modules available to the currently running kernel which makes life a bit hard if you needed hot plug support or something similar.
            pacman is so much nicer to interact with than apt because it doesn't do random magical stuff behind your back, and the packaging format itself is vastly simpler and much more transparent than debian.

            Comment


            • #26
              Originally posted by finalzone View Post
              Essentially that post just kicked a skeleton of dead horse.
              So called Unix principles is dead long time ago killed by X server among the first. Linux is not UNIX as a reminder as quoted below.
              Unix principles have nothing to do with UNIX, they are general guidelines for sane software development. The people postulating them happened to be working on Unix systems back then.

              Unix principles aren't dead, but yeah, X server and other stuff in Linux userspace did break them pretty bad and none broke a sweat. More modern concepts like systemd, Wayland are striving to follow Unix principles.

              Comment


              • #27
                I use some embedded ARM systems that run Linux and systemd (Fedora). They have a second kernel, but getting to it means connecting a USB-serial console so the prompt can be intercepted. Not exactly convenient, depending on the device. I've only needed to do this once, but it was sufficiently painful. Good feature.

                Comment


                • #28
                  Originally posted by starshipeleven View Post
                  Unix principles have nothing to do with UNIX, they are general guidelines for sane software development. The people postulating them happened to be working on Unix systems back then.

                  Unix principles aren't dead, but yeah, X server and other stuff in Linux userspace did break them pretty bad and none broke a sweat. More modern concepts like systemd, Wayland are striving to follow Unix principles.
                  Not to mention, far too many people are wedded to a specific implementation of them without ever actually looking into what they are.

                  See https://en.wikipedia.org/wiki/Unix_philosophy for more.

                  Heck, the main thing they tend to paraphrase accurately (use flat text files and design everything to consume and produce them) is itself an interpretation (published in 1994) of a statement which first appeared in 1978, "Expect the output of every program to become the input to another, as yet unknown, program. Don't clutter output with extraneous information. Avoid stringently columnar or binary input formats."

                  (Also, Note that it says "stringently columnar" rather than just "columnar" as in "don't make it difficult for programs to produce valid data by expecting exactly the right visual alignment"... something which various lightweight markup languages (eg. ReST) could learn from today for their table syntax.)
                  Last edited by ssokolow; 21 October 2018, 08:41 AM.

                  Comment


                  • #29
                    Or maybe I don't want systemd to touch or write that stupid counter to my fucking disk. When I go to recovery mode (which happens after the bootloader), there's a reason I want literally ZERO writes to my disk. This will make such thing impossible.

                    Pathetic feature.

                    Comment


                    • #30
                      Originally posted by Weasel View Post
                      Or maybe I don't want systemd to touch or write that stupid counter to my fucking disk. When I go to recovery mode (which happens after the bootloader), there's a reason I want literally ZERO writes to my disk. This will make such thing impossible.

                      Pathetic feature.
                      If it's writing to disk in recovery mode, then it's broken.

                      Heck, even outside of recovery mode, that sort of thing is what UEFI's support for storing small amounts of non-volatile OS-set metadata was intended for. (IIRC, Microsoft envisioned it being used for things like "Don't delay the boot with a prompt... just have a "reboot into BIOS menu" option in the OS.)

                      Comment

                      Working...
                      X