Open daniejstriata opened 6 months ago
+1 Now I can't connect to instance! ... Erasing : kernel-6.1.56-82.125.amzn2023.x86_64 239/250 warning: file /lib/modules/6.1.56-82.125.amzn2023.x86_64/modules.builtin.alias.bin: remove failed: No such file or directory
Cleanup : python3-pip-21.3.1-2.amzn2023.0.5.noarch 240/250 ... Running scriptlet: kernel-tools-6.1.61-85.141.amzn2023.x86_64 249/250 Cleanup : traceroute-3:2.1.0-13.amzn2023.0.2.x86_64 250/250 Running scriptlet: kernel-6.1.66-91.160.amzn2023.x86_64 250/250 warning: %posttrans(kernel-6.1.66-91.160.amzn2023.x86_64) scriptlet failed, exit status 1
Error in POSTTRANS scriptlet in rpm package kernel Running scriptlet: traceroute-3:2.1.0-13.amzn2023.0.2.x86_64 250/250
I'm curious why folks still do in place upgrades instead of building and testing a new image. What prevents you from using immutable updates?
I hit the same error but reinstalled the kernel-modules-extra after the upgrade and before the reboot. Seems to be stable so far.
I'm curious why folks still do in place upgrades instead of building and testing a new image. What prevents you from using immutable updates?
@bryantbiggs I'm testing in VirtualBox using KVM images and not sure if the issue I ran into would happen in EC2 as well. My EC2 hosts upgraded without problems. The software I support is not Cloud native or Open Source. It will always need to be installed by a Java based installer made for Windows, Linux and FreeBSD. Even if I can stop the service, unmount the data and configuration and mount that to the new Instance. I'll still need to install and enable the services as selected for the specific instance by rerunning the installer. The software is also not widely deployed so the efficiencies of running it any differently will not easily translate to Windows or FreeBSD instances running onprem.
It's a bug, sorry, the workaround for now is to remove the kernel-modules-extra
before the update, install the update, then re-install kernel-modules-extra
... we will fix it in an upcoming release.
Re-opening, we'll close it when the fix has been released
Describe the bug I was unable to upgrade to
2023.3.20231211
because of an error about kernel-modules-extra. I'm using a KVM image in VirtualBox.To Reproduce Steps to reproduce the behavior:
dnf upgrade --releasever=2023.3.20231211
Expected behavior upgrade should complete without errors.
Screenshots![image](https://github.com/amazonlinux/amazon-linux-2023/assets/13314239/92cfd5a0-b274-411c-a65c-56d73951b04f)
Desktop (please complete the following information): AL2023
Additional context I uninstalled kernel-modules-extra-6.1.61-85.141.amzn2023.x86_64 and was able to upgrade to 2023.3.20231211. On restarting the host (without adding kernel-modules-extra back) it is taking a long time booting in a loop:![image](https://github.com/amazonlinux/amazon-linux-2023/assets/13314239/4cbee9aa-e320-4743-87a4-718d23b9c688)