I’ve used a similar approach with the only difference being merge requests are deployed to staging repeatedly. When we want to release we then create a tag. Serves us well
@MnemonicCarrier
Жыл бұрын
Thanks - very interesting 😄I'm trying to set up a GitLab runner at the moment (so I don't have to use the slow shared runner).
@sergioteixeira3313
Жыл бұрын
when deploying to staging/prod in my opinion you should promote the ci/cd artifact/container, so you are sure what you tested in dev is what gets promoted.
@siimhs
Жыл бұрын
Indeed I was thinking similarly. Build once, deploy many. But I believe that this can be achieved with Kai's flow by skipping build stage and derive version to be deployed from commit or version tag.
@henrikholst7490
Жыл бұрын
Yes the methods can be combined. I do like the tag system especially with immutable prefix matching like this. However, for me, all in brunchless is best. No branches, no tags. Only tests with Automatic promotion all the way up.
@henrikholst7490
Жыл бұрын
If you are doing Kubernetes the promotion stuff becomes important. We should do a collab video on that. I'm not an influencer but I've got solid experience with this. I've iterated most things that doesn't work.
@kaihendry
Жыл бұрын
Happy to chat how you do it in Kubernetes! Email me or something.
@getrekt526
10 ай бұрын
Do I am the only one who see an animation on the like button while watching the video, at 9 minutes and 10 secondes, while he says "like the video" ?
@joelamoako6778
Жыл бұрын
make more real life devops videos. thanks for this though
@kaihendry
Жыл бұрын
This is real life 😂 Or give me an example of a video you want more of
@joelamoako6778
Жыл бұрын
@@kaihendry something like this, more of this, basically things we can actual use in our actual jobs in production
@puttenicole
Жыл бұрын
This trigger a new job every time you tag? Different artifacts?
@kaihendry
Жыл бұрын
Yes, different artifacts if you’re build isn’t reproducible indeed. It’s definitely not optimal I agree.
Пікірлер: 15