Winternewt for lightblue 1.1.0 deployment?

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Winternewt for lightblue 1.1.0 deployment?

jewzaam
Administrator
Sort of a internal discussion, but we'd likely make the playbooks part of lightblue.

Should we do it?  What are the gaps?  Copying Jen and Steve for awareness and comment.

Given we're looking at a 1.1.0 release soon (week or two?) I don't know that we have time, but maybe we could.  It would be nice to manage the roll-out with respect to multiple az's at least from winternewt instead of manually.
Reply | Threaded
Open this post in threaded view
|

Re: Winternewt for lightblue 1.1.0 deployment?

jkrieger
I'm up for the conversation if you guys are ready to have it.

A few things to note: we are still dependent on a 2nd ARB review (currently slated for 11/18) before we can give the green light on deploying anything to a production environment. But, we can use winternewt for all other environments.

We will also need to have a follow up conversation - with the request to decommission our command worker, we now have to white list commands. I'm worried that the functionality we discussed we had months ago will not be there now as a result. If your deployments are similar to the servicesjava deploys in SSE, you should be good to go.

We should also talk about the git worker specifically - this is what we can do now... http://release-engine.readthedocs.org/en/latest/worker-steps.html#git-cherrypickmerge

What else could you foresee needing?
Reply | Threaded
Open this post in threaded view
|

Re: Winternewt for lightblue 1.1.0 deployment?

jewzaam
Administrator
Sorry for taking so long to respond.  I am hoping just for the ability to update nodes in availability zones in sets.  There's no need to take anything out of rotation from a load balancing point of view and source code management is just a nice to have for a first pass.  So basically: 
* run puppet on first 1/2 of nodes
* run puppet on second 1/2 of nodes

Given the November 18 date looks like this won't be ready for when we plan on pushing the 1.1.0 release.  But we could try it out for dev if there's time.

On Wed, Oct 29, 2014 at 4:45 PM, jkrieger [via lightblue-dev] <[hidden email]> wrote:
I'm up for the conversation if you guys are ready to have it.

A few things to note: we are still dependent on a 2nd ARB review (currently slated for 11/18) before we can give the green light on deploying anything to a production environment. But, we can use winternewt for all other environments.

We will also need to have a follow up conversation - with the request to decommission our command worker, we now have to white list commands. I'm worried that the functionality we discussed we had months ago will not be there now as a result. If your deployments are similar to the servicesjava deploys in SSE, you should be good to go.

We should also talk about the git worker specifically - this is what we can do now... http://release-engine.readthedocs.org/en/latest/worker-steps.html#git-cherrypickmerge

What else could you foresee needing?


If you reply to this email, your message will be added to the discussion below:
http://dev.forum.lightblue.io/Winternewt-for-lightblue-1-1-0-deployment-tp205p206.html
To start a new topic under lightblue-dev, email [hidden email]
To unsubscribe from lightblue-dev, click here.
NAML