tinkerbell/smee

kernel arg `ip=dhcp` can cause very long boot time in Hook

Closed this issue · 1 comments

jacobweinstock commented

By default we are passing the kernel arg ip=dhcp in the auto.ipxe script. With machines with multiple network interfaces that might not all be cabled this has been shown to cause the kernel to try for quite long to get a dhcp response before giving up and moving on. This means it can take many minutes before HookOS is up and running.

I believe that was included by default as a carry over from the original OSIE.

Expected Behaviour

Current Behaviour

Possible Solution

I believe the original purpose was to make sure networking was up and running with OG Osie so that services that needed networking would function properly. With the move to Hook, we have a DHCP call as one of the onboot containers. We should be able to remove this ip=dhcp arg entirely from Boots. It will need to be test/validated.

Also, important to note is that removing this ip=dhcp doesn't mean a user can't add it back via --extra-kernel-args=ip=dhcp CLI flag.

Steps to Reproduce (for bugs)

Context

Your Environment

  • Operating System and version (e.g. Linux, Windows, MacOS):

  • How are you running Tinkerbell? Using Vagrant & VirtualBox, Vagrant & Libvirt, on Packet using Terraform, or give details:

  • Link to your project or a code example to reproduce issue:

jacobweinstock commented

resolved by #339