Encountered a situation, when userdata script did not start:

Jun 28 11:34:56 ip-10-0-96-47.eu-central-1.compute.internal systemd[1]: Reached target Multi-User System.
Jun 28 11:34:56 ip-10-0-96-47.eu-central-1.compute.internal systemd[1]: Starting Reconfigure the system from EC2 userdata on startup...
Jun 28 11:34:56 ip-10-0-96-47.eu-central-1.compute.internal amazon-init-start[978]: attempting to fetch configuration from EC2 user data...
Jun 28 11:34:56 ip-10-0-96-47.eu-central-1.compute.internal amazon-init-start[978]: /nix/store/gzx4ihb8hdxl1yj18kfhm5an6vvz59mn-unit-script-amazon-init-start/bin/amazon-init-start: line 16: /etc/ec2-metadata/user-data: No such file or directory
Jun 28 11:34:56 ip-10-0-96-47.eu-central-1.compute.internal amazon-init-start[978]: no user data is available
Jun 28 11:34:56 ip-10-0-96-47.eu-central-1.compute.internal systemd[1]: Finished Reconfigure the system from EC2 userdata on startup.

But with manualy reboot it worked:

Jun 28 11:40:34 ip-10-0-96-47.eu-central-1.compute.internal systemd[1]: Reached target Multi-User System.
Jun 28 11:40:34 ip-10-0-96-47.eu-central-1.compute.internal systemd[1]: Starting Reconfigure the system from EC2 userdata on startup...
Jun 28 11:40:34 ip-10-0-96-47.eu-central-1.compute.internal amazon-init-start[934]: attempting to fetch configuration from EC2 user data...
Jun 28 11:40:34 ip-10-0-96-47.eu-central-1.compute.internal amazon-init-start[934]: running EC2 User Data bootstrap

Encountered a situation, when userdata script did not start:

Jun 28 11:34:56 ip-10-0-96-47.eu-central-1.compute.internal systemd[1]: Reached target Multi-User System.
Jun 28 11:34:56 ip-10-0-96-47.eu-central-1.compute.internal systemd[1]: Starting Reconfigure the system from EC2 userdata on startup...
Jun 28 11:34:56 ip-10-0-96-47.eu-central-1.compute.internal amazon-init-start[978]: attempting to fetch configuration from EC2 user data...
Jun 28 11:34:56 ip-10-0-96-47.eu-central-1.compute.internal amazon-init-start[978]: /nix/store/gzx4ihb8hdxl1yj18kfhm5an6vvz59mn-unit-script-amazon-init-start/bin/amazon-init-start: line 16: /etc/ec2-metadata/user-data: No such file or directory
Jun 28 11:34:56 ip-10-0-96-47.eu-central-1.compute.internal amazon-init-start[978]: no user data is available
Jun 28 11:34:56 ip-10-0-96-47.eu-central-1.compute.internal systemd[1]: Finished Reconfigure the system from EC2 userdata on startup.

But with manual reboot it worked:

Jun 28 11:40:34 ip-10-0-96-47.eu-central-1.compute.internal systemd[1]: Reached target Multi-User System.
Jun 28 11:40:34 ip-10-0-96-47.eu-central-1.compute.internal systemd[1]: Starting Reconfigure the system from EC2 userdata on startup...
Jun 28 11:40:34 ip-10-0-96-47.eu-central-1.compute.internal amazon-init-start[934]: attempting to fetch configuration from EC2 user data...
Jun 28 11:40:34 ip-10-0-96-47.eu-central-1.compute.internal amazon-init-start[934]: running EC2 User Data bootstrap

Yes, I have IPv6 enabled:

# ifconfig
enX0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet 10.0.117.211  netmask 255.255.128.0  broadcast 10.0.127.255
        inet6 fe80::7b:b7ff:fe6c:ef85  prefixlen 64  scopeid 0x20<link>
        ether 02:7b:b7:6c:ef:85  txqueuelen 1000  (Ethernet)
        RX packets 15051562  bytes 2684395427 (2.5 GiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 9503900  bytes 7788542100 (7.2 GiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

That ifconfig suggests ipv6 is disabled

In reply to @arianvp:matrix.org

That ifconfig suggests ipv6 is disabled

Hello all! I've set up NixOS on AWS by booting into a NixOS AMI, then doing nixos-install onto a custom EBS volume. After that, I switched the boot device of my EC2 instance from the AMI-created temporary EBS volume to my new custom volume and booting the instance from it.

Now I have the AMI-created EBS volume still, although it's not attached to anything. I'd like to get rid of it. I think the cleanest way would be to detach the NixOS AMI from the EC2 instance. AFAIK I don't need the AMI anymore, since I have the OS installed to EBS. But I don't find a way to have an EC2 instance without an AMI!

Any advice on how to get rid of the AMI or its EBS volume?

Note that the AMI is and EBS volume already

e.g. if you’re using Terraform:

data "aws_ami" "nixos_arm64" {
  owners      = ["427812963091"]
  most_recent = true

  filter {
    name   = "name"
    values = ["nixos/24.05*"]
  }
  filter {
    name   = "architecture"
    values = ["arm64"] # or "x86_64"
  }
}

resource “aws_instance” “my_instance” {

  ami           = data.aws_ami.nixos_arm64.id
  instance_type = "t4g.nano"
  root_block_device {
    volume_size = 500
    volume_type = “io2”
  }

}

Will boot the NixOS AMI on an io2 EBS volume of 500GB size. NixOS will resize the partition and file system automatically on first boot