DevFW-CICD/ingress-nginx-helm
16
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Add mention about achieving zero-downtime rolling updates to Nginx Ingress Controller docs

Browse source
This commit is contained in:
tyranron 2017-06-27 19:12:02 +03:00
parent 1468fcb1aa
commit e16efc0a21
1 changed files with 9 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
controllers/nginx/README.md
View file

@ -480,6 +480,15 @@ I0316 12:24:37.610073 1 command.go:69] change in configuration detected. R
The NGINX ingress controller does not uses [Services](http://kubernetes.io/docs/user-guide/services) to route traffic to the pods. Instead it uses the Endpoints API in order to bypass [kube-proxy](http://kubernetes.io/docs/admin/kube-proxy/) to allow NGINX features like session affinity and custom load balancing algorithms. It also removes some overhead, such as conntrack entries for iptables DNAT. The NGINX ingress controller does not uses [Services](http://kubernetes.io/docs/user-guide/services) to route traffic to the pods. Instead it uses the Endpoints API in order to bypass [kube-proxy](http://kubernetes.io/docs/admin/kube-proxy/) to allow NGINX features like session affinity and custom load balancing algorithms. It also removes some overhead, such as conntrack entries for iptables DNAT.
__Note!__ NGINX ingress controller needs up to 10 seconds to rediscover new pods during rolling update. So, if you are using multi-pod deployment with rolling update strategy, make sure that [old pods are not removed before traffic is redirected to new ones](https://github.com/kubernetes/ingress/issues/322#issuecomment-298016539).
This can be achieved with simple `preStop` command in container spec (ensure that pod still serves well during this 15 secs before shutting down):
```yaml
lifecycle:
preStop:
exec:
command: ["sleep, "15"]
```
### NGINX notes ### NGINX notes