With the move to EBS Direct API uploads (PutSnapshotBlock), raw images
can be uploaded directly without conversion, making the qemu-img vpc
conversion step unnecessary.
This avoids a build-time format conversion and simplifies upload
workflows that use the EBS Direct API (e.g. NixOS/amis#364).
This kernel parameter was accidentally added to the image builder module
in 2018 (e4777ae2d8) when it should have been in the runtime config.
This causes a mismatch between AMIs built with the builder and systems
configured with just the runtime module.
Move it to modules/virtualisation/amazon-image.nix so that all EC2
instances get the recommended NVMe timeout, regardless of how they
were provisioned.
This prints the whole host public keys in the same format as cloud-init.
The main benefit is that this allows the direct generation of a known_hosts
file, simplifying client setup.
This commit also provides a new NixOS test that's an alternative to the
currently broken nixos/tests/ec2.nix tests.
vpc files use the extension "vhd". `make-disk-image-nix` contains a
lookup table, but does not expose that. vpc is the only format
supported by the amazon image which is affected. Format and extension
are the same for raw and qcow2.
Disk usage of the image has risen considerably since last release. We
still need to investigate the cause in detail. Increasing it for now
prevents build failures at least.
By default, `system.nixos.label` includes `system.nixos.tags` which
historically was not set for the amazon image.
Keeping the tag in the label of the generated disk image inside the
derivation (i.e. `image.filePath`) makes it easier to tell that the
file is for amazon (i.e.
`nixos-image-amazon-25.11pre-git-x86_64-linux.vpc`) while the "label"
in `image-info.json` for users of the AWS api stays stable (i.e.
`/25.05pre-git-x86_64-linux`).
`
these changes were generated with nixq 0.0.2, by running
nixq ">> lib.mdDoc[remove] Argument[keep]" --batchmode nixos/**.nix
nixq ">> mdDoc[remove] Argument[keep]" --batchmode nixos/**.nix
nixq ">> Inherit >> mdDoc[remove]" --batchmode nixos/**.nix
two mentions of the mdDoc function remain in nixos/, both of which
are inside of comments.
Since lib.mdDoc is already defined as just id, this commit is a no-op as
far as Nix (and the built manual) is concerned.
most of these are hidden because they're either part of a submodule that
doesn't have its type rendered (eg because the submodule type is used in
an either type) or because they are explicitly hidden. some of them are
merely hidden from nix-doc-munge by how their option is put together.
This is important since legacy bios mode is still the default for Intel
and AMD based instances on AWS. That is, even if your image is setup to
use UEFI on the OS level, the AMI will still use BIOS unless the boot
mode is explicitly set during registration.
Having a disks object with a dictionary of all the disks and their
properties makes it easier to process multi-disk images.
Note the rename of `label` to `system_label` is because `$label`i
is something of a special token to jq.
Introduce an AWS EC2 AMI which supports aarch64 and x86_64 with a ZFS
root.
This uses `make-zfs-image` which implies two EBS volumes are needed
inside EC2, one for boot, one for root. It should not matter which
is identified `xvda` and which is `xvdb`, though I have always
uploaded `boot` as `xvda`.
For reasons we haven't been able to work out, the aarch64 EC2 image now
regularly exceeds the output image size on hydra.nixos.org. As a
workaround, set this back to being statically sized again.
The other images do seem to build - it's just a case of the EC2 image
now being too large (occasionally non-determinstically).
As a temporary workaround for #120473 while the image builder is patched
to correctly look up disk sizes, partially revert
f3aa040bcb for EC2 disk images only.
We retain the type allowing "auto" but set the default back to the
previous value.
