github/kubernetes
1
0
Fork 0
mirror of https://github.com/k3s-io/kubernetes.git synced 2025-07-19 18:02:01 +00:00
Code Issues Projects Releases Wiki Activity

Include unschedulable nodes to compute dns replica

Browse Source 
This change picks up the new flag added to cluster autoscaler to include
unschedulable nodes when computing number of DNS replica. This fixes an
issue where cordoning large nodepools can cause a big drop in the number
of kube-dns replicas and cause existing workloads to see DNS outage.
This commit is contained in:
Pavithra Ramesh 2020-05-26 12:08:14 -07:00
parent 87eb688ec9
commit 10f300313e
1 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
cluster/addons/dns-horizontal-autoscaler/dns-horizontal-autoscaler.yaml
View File

@ -86,7 +86,7 @@ spec:
kubernetes.io/os: linux
containers:
- name: autoscaler
image: k8s.gcr.io/cluster-proportional-autoscaler-amd64:1.7.1
image: k8s.gcr.io/cluster-proportional-autoscaler-amd64:1.8.0
resources:
requests:
cpu: "20m"
@ -99,7 +99,7 @@ spec:
- --target={{.Target}}
# When cluster is using large nodes(with more cores), "coresPerReplica" should dominate.
# If using small nodes, "nodesPerReplica" should dominate.
- --default-params={"linear":{"coresPerReplica":256,"nodesPerReplica":16,"preventSinglePointFailure":true}}
- --default-params={"linear":{"coresPerReplica":256,"nodesPerReplica":16,"preventSinglePointFailure":true,"includeUnschedulableNodes":true}}
- --logtostderr=true
- --v=2
tolerations: