I think it is best to have a single person responsible for coordinating the containerization effort. Other epics can be created and assigned as needed, of course - but somebody needs to attend the meetings.
Feel free to assign yourself SE-1834, if you’re still up for it. We have a BTR meeting in an hour - it would be a good idea for you to attend, if you can.
Awesome, so glad to see SE-1834 become BB-4199 and be championed by @shimulch@giovannicimolin and @demid! Great too that you’ve got tasks to check out Tutorraform next sprint, so you can get a better idea of where you’d like to focus your efforts.
The original goal of this thread was to:
split the Ocim containerization epic into smaller projects, so that we can spread the capacity and learning required for implementing this major change across all of the cells.
But since it’s all Bebop people that have said they have time to work on this, I’m wondering if we still need to split up the work across cells?
Even if Bebop owns this, we can still:
Farm tasks out to interested people in other cells with capacity to spare.
Ocim has suffered from a problem of too many different developers working on it and taking it in different directions. Now is a good chance to reset and set a strong product vision for its future. So whatever process we pick, I’d like to ensure that someone (@shimulch with help from @giovannicimolin and @demid I guess :) ) is doing the overall planning and making sure that all the new work that’s done fits together well and builds toward the same end goal.
I’m thinking this week about how we can have more consistent leadership with our devops so will probably post separately about this too.
@braden, I am still trying to get my head around the responsibilities of the epic (since this is my first epic) and the design of Tutor and Tutorraform. I do agree in order to work towards a greater goal would require someone to take the responsibility to ensure everything fits together.
Since I don’t have any other epic/client to handle at the moment I can take that role with much-needed helps from @giovannicimolin and @demid as reviewers. I think this would be a new experience for me. And I will be glad to work on this epic as this seems to impact our dev ops greatly.
@demid, I think you have played with Tutorraform a bit with the learning ticket if I am not wrong. Do you think there is a need to split the epic?
Yep, @shimulch. I deployed instance to DO as was described in Braden’s tech talk. It’s great, though I have to learn a lot about K8S. Also, I’m going to use Tutor devstack for all development tickets I have in my current sprint to get used to it.
I think it’s better to do it when we know what changes Ocim and Tutorraform require.
What if we create a discovery ticket and try to create a prototype based on Ocim and Tutorraform, which would be able to deploy minimal containerized Open edX instances? Assignee of this ticket will gain a good understanding of what needs to be done for this epic. I would be happy to take such discovery, it’s going to be crazy fun.
I remember raising my hand for this, but I see it wasn’t in this thread
Anyway I’d love to get more involved on this, dockerization is something I am very interested in and Tutor looks amazing, I’d love to get the chance to contribute to it, so please count me in whatever needs to be done here
@jill Sorry for the late reply. @raul That’s really great. In this sprint I’ve created some discovery tickets BB-4337, BB-4339. And another one BB-4338 (in stretch goals). I would really love it if you want to review any one of those. Also since I could not take BB-4338 in this sprint, that task is also available for you to take.
Thanks @shimulch! I’ve suddenly ended up with other new epics and two mentees for next sprint though, so I shouldn’t take on anything else until I get an idea of how much extra work that’s going to generate. I don’t want to bite off more than I can chew!
BTW, we are still looking for an epic owner for sub epic BB-4400 - Building console Admin panel and improving maintainability of container-based instances.
This can start after we have achieved some milestones in BB-4398 and FAL-2081, when we have an MVP. But the project has a pretty tight deadline, so its better to find someone asap to start required discoveries.