packages/coredns-deployment
1
0
Fork 0
mirror of https://github.com/notherealmarco/coredns-deployment.git synced 2025-05-05 12:32:34 +02:00
Code Issues Projects Releases Packages 1 Wiki Activity Actions

preferred cause single node failure of coredns (#260)

Browse source 
preferredDuringSchedulingIgnoredDuringExecution cause single node failure in mostly deployment when increasing node one by one by cluster. In most deployment, when one node running coredns is lost, coredns service is outage at least 5mins.Suggest to use **requiredDuringSchedulingIgnoredDuringExecution** instead to avoid single node failure issue.

Signed-off-by: Eric <cu.eric.lee@gmail.com>
This commit is contained in:
Eric 2022-01-14 22:18:23 +08:00 • committed by GitHub
parent 22cdb6becb
commit e3a5468227
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 7 additions and 9 deletions
Download patch file Download diff file Expand all files Collapse all files

6
kubernetes/coredns.yaml.sed
View file

@ -105,10 +105,8 @@ spec:
kubernetes.io/os: linux kubernetes.io/os: linux
affinity: affinity:
podAntiAffinity: podAntiAffinity:
preferredDuringSchedulingIgnoredDuringExecution: requiredDuringSchedulingIgnoredDuringExecution:
- weight: 100 - labelSelector:
podAffinityTerm:
labelSelector:
matchExpressions: matchExpressions:
- key: k8s-app - key: k8s-app
operator: In operator: In