diff --git a/.build/images/dietpi-build b/.build/images/dietpi-build index 8f223bff03..ddf9b67d94 100755 --- a/.build/images/dietpi-build +++ b/.build/images/dietpi-build @@ -1253,22 +1253,6 @@ fi ####### Proxmox ############################ if [[ $VMTYPE =~ ^(proxmox|all)$ ]] then - # Enable systemd-logind for APCI functionality - # - Create loop device - FP_LOOP=$(losetup -f) - G_EXEC losetup "$FP_LOOP" "$OUTPUT_IMG_NAME.img" - G_EXEC partprobe "$FP_LOOP" - G_EXEC partx -u "$FP_LOOP" - # - Mount and edit dietpi.txt - G_EXEC mkdir rootfs - G_EXEC mount "${FP_LOOP}p1" rootfs - # - Remove loop device - G_EXEC sync - G_EXEC sleep 1 - G_EXEC umount -R rootfs - G_EXEC rmdir rootfs - G_EXEC losetup -d "$FP_LOOP" - # Convert raw image to QCOW2 image_name=${OUTPUT_IMG_NAME/_VM-/_Proxmox-} G_EXEC qemu-img convert -c -O qcow2 "$OUTPUT_IMG_NAME.img" "$image_name.qcow2" diff --git a/CHANGELOG.txt b/CHANGELOG.txt index 76b2e25ae6..76c63b86c1 100644 --- a/CHANGELOG.txt +++ b/CHANGELOG.txt @@ -7,6 +7,7 @@ New images: - NanoPi M6 | Support for this FriendlyELEC SBC with RK3588 SoC has been added to DietPi. Enhancements: +- Virtual Machines | systemd-logind is now enabled by default on VMs, as it provides ACPI functionality, needed to properly shutdown or reboot the VM from the virtualiser software, usually expected by users. - DietPi-DDNS | The "IPv6or4" option to update IPv6 only, if supported by server, network and provider, has been replaced with "IPv4and6". A server being reachable via IPv6 only is rarely wanted, as many networks do not support it. Instead, usually one will want to have it reachable via both, IPv4 as well as IPv6, which is now possible when using DietPi-DDNS, and the new default. If, e.g. for security reason, IPv6 only is wanted, this option of course remains available, like before. Many thanks to @LOGIN-TB for doing this suggestion: https://github.com/MichaIng/DietPi/issues/7278 - DietPi-DDNS | The cron job does now log server response messages and connection errors separately with respective severities. Some DDNS providers do not return an HTTP error code, but an error text embedded into a regular HTTP 200 response. This, as well as success responses can now be seen via "journalctl -t dietpi-ddns". Many thanks to @jtmoon79 for doing this suggestion: https://github.com/MichaIng/DietPi/issues/5954 - DietPi-Services | "dietpi-services start" will not start disabled services anymore. This aligns with the behaviour of the "restart" command, which as well skips disabled services. The script is used within other DietPi scripts to (re)start services after maintenance operations, and it is unexpected when services, who were not running before, but explicitly disabled, are running afterwards. To manually start/stop individual services from the console, we recommend using "systemctl" directly, like on any other Linux distribution with systemd. Many thanks to @intiplink for reporting this unexpected behaviour while using dietpi-drive_manager: https://github.com/MichaIng/DietPi/issues/7302