equivalent of "up" "down" ? about vagrant-openshift

[danmcp]

@jayunit100 I don't think any of the commands you ran actually started openshift. You could run:

vagrant ssh -c "sudo systemctl start openshift"

[jayunit100]

@danmcp - should we simply have a vagrant openshift-up similar to the upstream kube-up ? cc @smarterclayton .

• builds the source
• deploys it
• starts the actual services with working DNS, metrics, etc.
• env var for number of nodes

in another upstream issue (openshift/origin#6340), i think we had a similar conversation, and it mostly ended with the idea that there was not much value in specifying the origin vagrant recipes.

Im thinking maybe vagrant-openshift is a better home for a universal developer's openshift ? I could be wrong though...

[danmcp]

@jayunit100 The issue is that the devenvs are used for 2 purposes (development and testing). If you use them for development, you don't want the services started because they conflict with running tests. If you use them for testing you do want them started. The default OOTB is to favor the development scenario and not start everything up. There are other VMs that exist (The CDK and the one @thesteve0 and @jwhonce created) that are meant to handle the end user scenarios. I don't personally have any objection to making vagrant-openshift better within its niche, but I don't want to add even another option to the list that everyone has to choose between.

[jayunit100]

hmmm ok, so, to be clear: you're saying that building, and then running an openshift cluster from source is actually not a use case that vagrant-openshift is meant to cover? if so i guess i should close this issue (but maybe im missing something)

[danmcp]

@jayunit100 It's just not its primary use case. And the more we do to make that its use case, the more it will conflict with either the CDK or openshift-ansible. The momentum seems to be more in the other direction of making openshift-ansible more accessible for these types of scenarios.

[smarterclayton]

Ideally in the origin repo we'd have cluster/up.sh or similar that depends
on a known commit of openshift-ansible.

On Tue, Mar 22, 2016 at 8:24 AM, jay vyas [email protected] wrote:

Closed #420 #420.

—
You are receiving this because you were mentioned.
Reply to this email directly or view it on GitHub
#420 (comment)

[jayunit100]

Yeah, It seems to me that Such a recipe would obviate the need for any other tooling for the majority of dev AND test use cases:.... After all, upstream kube can be entirely developed using nothing other than vim and kube-up.sh,