From a13f82a2bb79e78bcc76c691c744ce852a70bdeb Mon Sep 17 00:00:00 2001 From: Marcin Wielgus Date: Mon, 4 Sep 2017 19:54:36 +0200 Subject: [PATCH] Graceful termination update in CA FAQ --- cluster-autoscaler/FAQ.md | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/cluster-autoscaler/FAQ.md b/cluster-autoscaler/FAQ.md index f6fcde4db1..d2502568a7 100644 --- a/cluster-autoscaler/FAQ.md +++ b/cluster-autoscaler/FAQ.md @@ -259,7 +259,9 @@ From 0.5 CA (K8S 1.6) respects PDB. Before starting to delete a node CA makes su ### Does CA respect GracefulTermination in scale down? -CA gives pods at most 1 min graceful termination time. If the pod is not stopped within this 1 min the node is deleted anyway. +CA, from version 0.7, gives pods at most 10 min graceful termination time. If the pod is not stopped within +these 10 min then the node is deleted anyway. Earlier versions of CA gave 1 min or didn't respect graceful +termination at all. ### How does CA deal with unready nodes in version <= 0.4.0?