@Kensan Nope, we didn't - in practice most MirageOS DHCP clients are already on their own vlan/subnet so it hasn't been called for, and it makes the lease acquisition (and therefore time to real unikernel start) slower. It wouldn't be hard to implement though.
@Kensan If we follow through on the work juga0 started at the hack retreat, we'll probably have to do something about this since we'll have a UNIX-usable charrua-based client which will probably be interacting in messy real-world scenarios where IP conflicts might happen :/
@yomimono Not necessarily: our DHCP client (ADHCP) has been in production use for several years now and we have not implemented ACD update until now. I have also been using it on my systems and I have not had any issues in that regard...
@yomimono I've now glanced at this toot several times and each time I've misread it as saying that most MirageOS deployments have their own villain 😆
@jamey it me :P
@yomimono Thanks for the insight, makes sense. The added delay is actually what I am concerned about. The values in the RFC add several seconds so I was wondering if there are values used “in the field” which are more sensible given today’s networks... Thanks again!