> [0]
>
> Is anything using launch_time? We removed it from the EC2 machines because
> nothing was using it (the only use was in ZookeeperConnect, and niemeyer
> opined it was costing more complexity than it was worth:
> https://code.launchpad.net/~fwereade/ensemble/cobbler-zk-
> connect/+merge/71734).
thanks, that's good to know.
>
> [1]
>
> Please add a test for .private_dns_name (it needs to exist for us to generate
> the zookeeper hosts)
>
it would be if local-dev was using that pathway.. easy to add in though, but it would be unused.
> [2]
>
> ...finally I guess I'm a bit confused about the hostname anyway. Will
> "localhost" resolve to the right place, even from within a container? If so,
> fine (but, eww!); if not, surely we should have something that does resolve to
> the right place?
>
> I don't feel it's quite approvable as-is, but ping me to discuss.
the containers don't use this to resolve to the host, this is soley to satisfy existing constraint notions of machine usage and assignment.
> [0] /code.launchpad .net/~fwereade/ ensemble/ cobbler- zk- +merge/ 71734).
>
> Is anything using launch_time? We removed it from the EC2 machines because
> nothing was using it (the only use was in ZookeeperConnect, and niemeyer
> opined it was costing more complexity than it was worth:
> https:/
> connect/
thanks, that's good to know.
>
> [1]
>
> Please add a test for .private_dns_name (it needs to exist for us to generate
> the zookeeper hosts)
>
it would be if local-dev was using that pathway.. easy to add in though, but it would be unused.
> [2]
>
> ...finally I guess I'm a bit confused about the hostname anyway. Will
> "localhost" resolve to the right place, even from within a container? If so,
> fine (but, eww!); if not, surely we should have something that does resolve to
> the right place?
>
> I don't feel it's quite approvable as-is, but ping me to discuss.
the containers don't use this to resolve to the host, this is soley to satisfy existing constraint notions of machine usage and assignment.