On 2 November 2017 at 07:16, Mark Shuttleworth <m...@ubuntu.com> wrote:
> On 11/02/2017 04:56 AM, Chris Lee wrote:
>
> A new development release of Juju is here, 2.3-beta2.
>
>
> 2.3 is looking great, and is worth a test run for those of you with larger
> models and an interest in cross-model relations.
>
>
> ## New and Improved
>
> * Deploying bundles can now target existing machines
>
> juju deploy some-bundle --use-existing-machines --bundle-machine 3=4
>
>
> From a D-R-Y perspective, we --dont --need --all --those --options :)
>
> Perhaps just:
>
> juju deploy --map-machines A=B,C=D
>
> ... or some variant of that?
>
> Let's use the betas to refine and condense and clarify.
+1 to that. I'm wondering if use-existing-machines is ever appropriate
on its own, as the machine numbers in a model are ephemeral but
machine numbers in a bundle are static.
ISTM that there's a strong connection between specifying the machines that
a bundle is to be run on and specifying the machine that a newly deployed
service will run on. Going with that thought, how about
juju deploy --to 3=0,4=3
and losing the --use-existing-machines flag? That way there are
no new flags at all for this feature and the intent of existing flags
is still preserved.
rog.
--
Juju-dev mailing list
Juju-dev@lists.ubuntu.com
Modify settings or unsubscribe at:
https://lists.ubuntu.com/mailman/listinfo/juju-dev
