From faaae90b41085a1e1555e0b474c3e0548b10100d Mon Sep 17 00:00:00 2001 From: adriananeci Date: Fri, 21 Feb 2025 16:34:40 +0200 Subject: [PATCH] Update content/en/docs/concepts/workloads/controllers/deployment.md Co-authored-by: Tim Bannister --- .../concepts/workloads/controllers/deployment.md | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/content/en/docs/concepts/workloads/controllers/deployment.md b/content/en/docs/concepts/workloads/controllers/deployment.md index ae55d82d7c..5dfe164055 100644 --- a/content/en/docs/concepts/workloads/controllers/deployment.md +++ b/content/en/docs/concepts/workloads/controllers/deployment.md @@ -1079,13 +1079,15 @@ Explicitly setting this field to 0, will result in cleaning up all the history o thus that Deployment will not be able to roll back. {{< /note >}} -{{< note >}} -The cleanup will ONLY start after a deployment reaches a -[complete state](/docs/concepts/workloads/controllers/deployment/#complete-deployment) -For example, if pods are crash looping, and there are multiple rolling updates +The cleanup only starts **after** a Deployment reaches a +[complete state](/docs/concepts/workloads/controllers/deployment/#complete-deployment). +If you set `.spec.revisionHistoryLimit` to 0, any rollout nonetheless triggers creation of a new +ReplicaSet before Kubernetes removes the old one. + +Even with a non-zero revision history limit, you can have more ReplicaSets than the limit +you configure. For example, if pods are crash looping, and there are multiple rolling updates events triggered over time, you might end up with more ReplicaSets than the -`.spec.revisionHistoryLimit` because the deployment never reaches a complete state. -{{< /note >}} +`.spec.revisionHistoryLimit` because the Deployment never reaches a complete state. ## Canary Deployment