Skip to content

Commit 887d371

Browse files
committed
add details for becoming a PRR approver
1 parent e3317ab commit 887d371

File tree

1 file changed

+22
-3
lines changed

1 file changed

+22
-3
lines changed

sig-architecture/production-readiness.md

Lines changed: 22 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -80,9 +80,28 @@ To become a reviewer:
8080
* Read/study previous PRR comments and production readiness responses in existing KEPs.
8181
* Choose some KEPs requiring PRR and perform a review. Put "shadow prod readiness review"
8282
in your review comments so that the assigned PRR approver knows your intent.
83-
* After at least one release cycle, if you have shown good judgement and quality reviews,
84-
you can propose yourself as approver by submitting a PRR to add your GitHub
85-
handle to the `prod-readiness-approvers` alias in [OWNERS_ALIASES].
83+
84+
### Becoming an approver
85+
86+
After serving as reviewer/shadow for at least one release and showing good judgement and quality reviews,
87+
you can propose yourself as an approver by submitting a PR to add your GitHub
88+
handle to the `prod-readiness-approvers` alias in [OWNERS_ALIASES].
89+
90+
When submitting the PR, you should include references to KEPs you reviewed that demonstrated a good variety
91+
of different situations.
92+
Here is a good starting point (remember that one PR can cover multiple categories):
93+
94+
* Transitions from new to alpha
95+
* Transitions from alpha to beta
96+
* Transitions from beta to GA
97+
* Must have successfully reviewed at least three enhancements that require coordination between multiple components.
98+
* Must have successfully reviewed at least three enhancements that require version skew consideration (both HA and component skew):
99+
does behavior fail safely and eventually reconcile.
100+
* Must have successfully reviewed at least three enhancements that are outside your primary domain.
101+
* Examples where the feature requires considering the case of administering thousands of clusters.
102+
This comes up frequently for host-based features in storage, node, or networking.
103+
* Examples where the feature requires considering the case of very large clusters. This is commonly covered by metrics.
104+
86105

87106
## Finding KEPs needing prod readiness review
88107

0 commit comments

Comments
 (0)