Deploy failure tolerance knobs

[davidreuss]

Describe the problem/challenge you have

For certain deploys your tolerance will be low for container restarts. It should be possible to exit out of the deploy early, but also include configurable behaviour to adjust timeout/tolerance.

Describe the solution you'd like Similar to wait behaviours there should be health criteria for kapp to honor. For instance the number of restarts tolerated for containers in a pod.

Anything else you would like to add: kapp is great, easy to get going. I'm happy about how it works on many levels.

Slack thread for context:

https://kubernetes.slack.com/archives/CH8KCCKA5/p1635440293215900

---

Vote on this request

This is an invitation to the community to vote on issues, to help us prioritize our backlog. Use the "smiley face" up to the right of this comment to vote.

👍 "I would like to see this addressed as soon as possible" 👎 "There are other more important things to focus on right now"

We are also happy to receive and review Pull Requests if you want to help working on this issue.

[renuy]

Hey @davidreuss, first off thanks for the kind words. this keeps us motivated. :)

This looks like a fair ask. Will triage this issue, and keep you posted here.

[rohitagg2020]

Hi @davidreuss ,

I have gone through the slack discussion and the issue description. I need certain clarifications. Please share your thoughts on the same:

1. Are you looking for exit based on timeout. If yes, we already have something like this: --wait-resource-timeout duration Maximum amount of time to wait for a resource in wait phase (0s means no timeout)

e.g. kapp deploy -a appName -f fileName.yaml --wait-resource-timeout 2m In this case, the kapp deploy will exit if any resource takes > 2 min to reach its desired state.

2. Are you looking for exit based on number of container restarts. As restarts are only for the pods, do you have any other resources and their respective conditions in mind.