diff --git a/pages/elastic-metal/troubleshooting/index.mdx b/pages/elastic-metal/troubleshooting/index.mdx index 537594955d..153d2059dc 100644 --- a/pages/elastic-metal/troubleshooting/index.mdx +++ b/pages/elastic-metal/troubleshooting/index.mdx @@ -66,4 +66,5 @@ productIcon: ElasticMetalProductIcon - [Troubleshooting remote access (KVM over IP) issues on Elastic Metal servers](/elastic-metal/troubleshooting/troubleshoot-remote-access-issues/) - [Troubleshooting SSH connection issues on Elastic Metal servers](/elastic-metal/troubleshooting/troubleshoot-ssh-issues/) - [Installation from virtual media not possible on A610R/A215R Elastic Metal servers](/elastic-metal/troubleshooting/cant-install-from-virtualmedia/) + - [Troubleshooting Error 500 on boot after custom installation on an Elastic Metal server](/elastic-metal/troubleshooting/troubleshoot-error-500-on-boot/) diff --git a/pages/elastic-metal/troubleshooting/troubleshoot-error-500-on-boot.mdx b/pages/elastic-metal/troubleshooting/troubleshoot-error-500-on-boot.mdx new file mode 100644 index 0000000000..9e480b5913 --- /dev/null +++ b/pages/elastic-metal/troubleshooting/troubleshoot-error-500-on-boot.mdx @@ -0,0 +1,62 @@ +--- +title: Troubleshooting Error 500 on boot after custom installation on an Elastic Metal server +description: Solutions for resolving boot issues after a custom installation on an Elastic Metal server. +tags: elastic-metal troubleshooting installation +dates: + validation: 2025-12-05 + posted: 2025-12-05 +--- + +## Problem +Your Scaleway Elastic Metal server is blocked during the boot process with an iPXE *Error 500* after performing a custom installation. + +### Symptoms +- The server jumps directly into iPXE mode even though a bootable disk appears to be present. +- The console displays repeated attempts to PXE boot, ending with an `iPXE 500` error. +- After several failed attempts, the system may eventually boot if the correct EFI entry exists. + +## Possible causes +- Incorrect or missing UEFI boot entries: The OS installer did not create (or could not create) the necessary UEFI NVRAM entry for the new system. +- EFI bootloader installed in a non-standard location: The EFI binary (e.g., `grubx64.efi`, `shimx64.efi`, `bootmgfw.efi`) is not located where the Bootscript expects it. +- No generic fallback loader (`\EFI\Boot\bootx64.efi`): Some installers rely solely on vendor-specific paths and do not populate the fallback bootloader that UEFI firmware attempts automatically. +- Incomplete OS installation: The bootloader files may be missing or corrupted if the installation did not finish properly. +- Overwritten or erased NVRAM entries: Some custom installers clear existing UEFI entries but fail to add new ones. +- Bootscript fallback behavior: Scaleway’s Bootscript attempts PXE boot when it does not find a valid EFI loader. Since it cannot predict the bootloader path for custom installations, PXE will fail with a 500 error by design. + +## Solution + +On Elastic Metal servers using UEFI, the PXE boot mechanism depends on the system’s ability to locate a valid EFI bootloader on disk. +For custom installations, bootscript cannot know in advance which EFI binary to use, because the path varies depending on the operating system you install. +For example: + +- Windows uses: `\EFI\Microsoft\Boot\bootmgfw.efi` +- Many Linux distributions use: `\EFI\{distro}\grubx64.efi` +- Generic fallback loader (if present): `\EFI\Boot\bootx64.efi` + +Because of this uncertainty: + +1. The Bootscript attempts to identify a valid EFI loader. +2. It fails to locate one (because the path is not predictable). +3. It falls back to PXE boot, which fails with **iPXE Error 500**. +4. The server then moves on to the next boot option. + +This behavior is expected and does **not** mean that your hardware or installation is broken. + +If your custom OS installation completed correctly, it should have created its own boot entry (e.g., `\EFI\Microsoft\Boot\bootmgfw.efi` or a GRUB entry). Once that entry exists, the server will use it and boot normally after the PXE fallback. + +### What you should do +- Ensure your custom installation places a valid EFI loader on the disk. +- Verify that the EFI files exist in the partition mounted at `/boot/efi` (Linux) or the EFI System Partition (ESP). +- Confirm that a UEFI boot entry was created by the installer. + - On Linux, you can check with `efibootmgr -v`. +- If necessary, manually create or restore the EFI loader in: +```bash +\EFI\Boot\bootx64.efi +``` +(the generic fallback path recognized by most UEFI firmwares). + + +This issue is a direct result of a custom installation that does not properly declare or place the EFI bootloader and is not caused by Scaleway. +Since the installation is custom, determining the correct EFI paths and entries is your responsibility. +The iPXE 500 error is simply a fallback mechanism and can be safely ignored once a correct boot entry is present. + \ No newline at end of file