Replace RC with Deployments
Janet Kuo
parent
1972537aef
commit
7077aac069
|
|
@ -7,107 +7,91 @@
|
|||
|
||||
## Launching a set of replicas using a configuration file
|
||||
|
||||
Kubernetes creates and manages sets of replicated containers (actually, replicated [Pods](/docs/user-guide/pods)) using [*Replication Controllers*](/docs/user-guide/replication-controller).
|
||||
Kubernetes creates and manages sets of replicated containers (actually, replicated [Pods](/docs/user-guide/pods)) using [*Deployments*](/docs/user-guide/deployments).
|
||||
|
||||
A replication controller simply ensures that a specified number of pod "replicas" are running at any one time. If there are too many, it will kill some. If there are too few, it will start more. It's analogous to Google Compute Engine's [Instance Group Manager](https://cloud.google.com/compute/docs/instance-groups/manager/) or AWS's [Auto-scaling Group](http://docs.aws.amazon.com/AutoScaling/latest/DeveloperGuide/AutoScalingGroup) (with no scaling policies).
|
||||
A Deployment simply ensures that a specified number of pod "replicas" are running at any one time. If there are too many, it will kill some. If there are too few, it will start more. It's analogous to Google Compute Engine's [Instance Group Manager](https://cloud.google.com/compute/docs/instance-groups/manager/) or AWS's [Auto-scaling Group](http://docs.aws.amazon.com/AutoScaling/latest/DeveloperGuide/AutoScalingGroup.html) (with no scaling policies).
|
||||
|
||||
The replication controller created to run nginx by `kubectl run` in the [Quick start](/docs/user-guide/quick-start) could be specified using YAML as follows:
|
||||
The Deployment created to run nginx by `kubectl run` in the [Quick start](/docs/user-guide/quick-start) could be specified using YAML as follows:
|
||||
|
||||
```yaml
|
||||
apiVersion: v1
|
||||
kind: ReplicationController
|
||||
metadata:
|
||||
name: my-nginx
|
||||
spec:
|
||||
replicas: 2
|
||||
template:
|
||||
metadata:
|
||||
labels:
|
||||
app: nginx
|
||||
spec:
|
||||
containers:
|
||||
- name: nginx
|
||||
image: nginx
|
||||
ports:
|
||||
- containerPort: 80
|
||||
```
|
||||
{% include code.html language="yaml" file="run-my-nginx.yaml" ghlink="/docs/user-guide/run-my-nginx.yaml" %}
|
||||
|
||||
Some differences compared to specifying just a pod are that the `kind` is `ReplicationController`, the number of `replicas` desired is specified, and the pod specification is under the `template` field. The names of the pods don't need to be specified explicitly because they are generated from the name of the replication controller.
|
||||
View the [replication controller API
|
||||
object](/docs/api-reference/v1/definitions/#_v1_replicationcontroller)
|
||||
Some differences compared to specifying just a pod are that the `kind` is `Deployment`, the number of `replicas` desired is specified, and the pod specification is under the `template` field. The names of the pods don't need to be specified explicitly because they are generated from the name of the Deployment.
|
||||
View the [Deployment API
|
||||
object](/docs/api-reference/extensions/v1beta1/definitions/#_v1beta1_deployment)
|
||||
to view the list of supported fields.
|
||||
|
||||
This replication controller can be created using `create`, just as with pods:
|
||||
This Deployment can be created using `create`, just as with pods:
|
||||
|
||||
```shell
|
||||
$ kubectl create -f ./nginx-rc.yaml
|
||||
replicationcontrollers/my-nginx
|
||||
$ kubectl create -f ./run-my-nginx.yaml
|
||||
deployment "my-nginx" created
|
||||
```
|
||||
|
||||
Unlike in the case where you directly create pods, a replication controller replaces pods that are deleted or terminated for any reason, such as in the case of node failure. For this reason, we recommend that you use a replication controller for a continuously running application even if your application requires only a single pod, in which case you can omit `replicas` and it will default to a single replica.
|
||||
Unlike in the case where you directly create pods, a Deployment replaces pods that are deleted or terminated for any reason, such as in the case of node failure. For this reason, we recommend that you use a Deployment for a continuously running application even if your application requires only a single pod, in which case you can omit `replicas` and it will default to a single replica.
|
||||
|
||||
## Viewing replication controller status
|
||||
## Viewing Deployment status
|
||||
|
||||
You can view the replication controller you created using `get`:
|
||||
You can view the Deployment you created using `get`:
|
||||
|
||||
```shell
|
||||
$ kubectl get rc
|
||||
CONTROLLER CONTAINER(S) IMAGE(S) SELECTOR REPLICAS
|
||||
my-nginx nginx nginx app=nginx 2
|
||||
$ kubectl get deployment
|
||||
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE
|
||||
my-nginx 2 2 2 2 6s
|
||||
```
|
||||
|
||||
This tells you that your controller will ensure that you have two nginx replicas.
|
||||
This tells you that your Deployment will ensure that you have two nginx replicas (desired replicas = 2).
|
||||
|
||||
You can see those replicas using `get`, just as with pods you created directly:
|
||||
|
||||
```shell
|
||||
$ kubectl get pods
|
||||
NAME READY STATUS RESTARTS AGE
|
||||
my-nginx-065jq 1/1 Running 0 51s
|
||||
my-nginx-buaiq 1/1 Running 0 51s
|
||||
NAME READY STATUS RESTARTS AGE
|
||||
my-nginx-3800858182-9hk43 1/1 Running 0 8m
|
||||
my-nginx-3800858182-e529s 1/1 Running 0 8m
|
||||
```
|
||||
|
||||
## Deleting replication controllers
|
||||
## Deleting Deployments
|
||||
|
||||
When you want to kill your application, delete your replication controller, as in the [Quick start](/docs/user-guide/quick-start):
|
||||
When you want to kill your application, delete your Deployment, as in the [Quick start](/docs/user-guide/quick-start):
|
||||
|
||||
```shell
|
||||
$ kubectl delete rc my-nginx
|
||||
replicationcontrollers/my-nginx
|
||||
$ kubectl delete deployment/my-nginx
|
||||
deployment "my-nginx" deleted
|
||||
```
|
||||
|
||||
By default, this will also cause the pods managed by the replication controller to be deleted. If there were a large number of pods, this may take a while to complete. If you want to leave the pods running, specify `--cascade=false`.
|
||||
By default, this will also cause the pods managed by the Deployment to be deleted. If there were a large number of pods, this may take a while to complete. If you want to leave the pods running instead, specify `--cascade=false`.
|
||||
|
||||
If you try to delete the pods before deleting the replication controller, it will just replace them, as it is supposed to do.
|
||||
If you try to delete the pods before deleting the Deployments, it will just replace them, as it is supposed to do.
|
||||
|
||||
## Labels
|
||||
|
||||
Kubernetes uses user-defined key-value attributes called [*labels*](/docs/user-guide/labels) to categorize and identify sets of resources, such as pods and replication controllers. The example above specified a single label in the pod template, with key `app` and value `nginx`. All pods created carry that label, which can be viewed using `-L`:
|
||||
Kubernetes uses user-defined key-value attributes called [*labels*](/docs/user-guide/labels) to categorize and identify sets of resources, such as pods and Deployments. The example above specified a single label in the pod template, with key `run` and value `my-nginx`. All pods created carry that label, which can be viewed using `-L`:
|
||||
|
||||
```shell
|
||||
$ kubectl get pods -L app
|
||||
NAME READY STATUS RESTARTS AGE APP
|
||||
my-nginx-afv12 0/1 Running 0 3s nginx
|
||||
my-nginx-lg99z 0/1 Running 0 3s nginx
|
||||
$ kubectl get pods -L run
|
||||
NAME READY STATUS RESTARTS AGE RUN
|
||||
my-nginx-3800858182-1v53o 1/1 Running 0 46s my-nginx
|
||||
my-nginx-3800858182-2ds1q 1/1 Running 0 46s my-nginx
|
||||
```
|
||||
|
||||
The labels from the pod template are copied to the replication controller's labels by default, as well -- all resources in Kubernetes support labels:
|
||||
The labels from the pod template are copied to the Deployment's labels by default, as well -- all resources in Kubernetes support labels:
|
||||
|
||||
```shell
|
||||
$ kubectl get rc my-nginx -L app
|
||||
CONTROLLER CONTAINER(S) IMAGE(S) SELECTOR REPLICAS APP
|
||||
my-nginx nginx nginx app=nginx 2 nginx
|
||||
$ kubectl get deployment/my-nginx -L run
|
||||
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE RUN
|
||||
my-nginx 2 2 2 2 2m my-nginx
|
||||
```
|
||||
|
||||
More importantly, the pod template's labels are used to create a [`selector`](/docs/user-guide/labels/#label-selectors) that will match pods carrying those labels. You can see this field by requesting it using the [Go template output format of `kubectl get`](/docs/user-guide/kubectl/kubectl_get):
|
||||
|
||||
```shell
|
||||
$ kubectl get rc my-nginx -o template --template="{{.spec.selector}}"
|
||||
map[app:nginx]
|
||||
$ kubectl get deployment/my-nginx -o template --template="{{.spec.selector}}"
|
||||
map[matchLabels:map[run:my-nginx]]
|
||||
```
|
||||
|
||||
You could also specify the `selector` explicitly, such as if you wanted to specify labels in the pod template that you didn't want to select on, but you should ensure that the selector will match the labels of the pods created from the pod template, and that it won't match pods created by other replication controllers. The most straightforward way to ensure the latter is to create a unique label value for the replication controller, and to specify it in both the pod template's labels and in the selector.
|
||||
You could also specify the `selector` explicitly, such as if you wanted to specify labels in the pod template that you didn't want to select on, but you should ensure that the selector will match the labels of the pods created from the pod template, and that it won't match pods created by other Deployments. The most straightforward way to ensure the latter is to create a unique label value for the Deployment, and to specify it in both the pod template's labels and in the selector's
|
||||
matchLabels.
|
||||
|
||||
## What's next?
|
||||
|
||||
[Learn about exposing applications to users and clients, and connecting tiers of your application together.](/docs/user-guide/connecting-applications)
|
||||
[Learn about exposing applications to users and clients, and connecting tiers of your application together.](/docs/user-guide/connecting-applications)
|
||||
|
|
|
|||
|
|
@ -6,44 +6,36 @@ This guide will help you get oriented to Kubernetes and running your first conta
|
|||
* TOC
|
||||
{:toc}
|
||||
|
||||
## Launching a simple application
|
||||
## Launching a simple application, and exposing it to the Internet
|
||||
|
||||
Once your application is packaged into a container and pushed to an image registry, you're ready to deploy it to Kubernetes.
|
||||
Through integration with some cloud providers (for example Google Compute Engine and AWS EC2), Kubernetes also enables you to request it to provision a public IP address for your application.
|
||||
|
||||
For example, [nginx](http://wiki.nginx.org/Main) is a popular HTTP server, with a [pre-built container on Docker hub](https://registry.hub.docker.com/_/nginx/). The [`kubectl run`](/docs/user-guide/kubectl/kubectl_run) command below will create two nginx replicas, listening on port 80.
|
||||
For example, [nginx](http://wiki.nginx.org/Main) is a popular HTTP server, with a [pre-built container on Docker hub](https://registry.hub.docker.com/_/nginx/). The [`kubectl run`](/docs/user-guide/kubectl/kubectl_run) command below will create two nginx replicas, listening on port 80, and a public IP address for your application.
|
||||
|
||||
```shell
|
||||
$ kubectl run my-nginx --image=nginx --replicas=2 --port=80
|
||||
CONTROLLER CONTAINER(S) IMAGE(S) SELECTOR REPLICAS
|
||||
my-nginx my-nginx nginx run=my-nginx 2
|
||||
$ kubectl run my-nginx --image=nginx --replicas=2 --port=80 --expose --service-overrides='{ "spec": { "type": "LoadBalancer" } }'
|
||||
service "my-nginx" created
|
||||
deployment "my-nginx" created
|
||||
```
|
||||
|
||||
You can see that they are running by:
|
||||
|
||||
```shell
|
||||
$ kubectl get po
|
||||
NAME READY STATUS RESTARTS AGE
|
||||
my-nginx-l8n3i 1/1 Running 0 29m
|
||||
my-nginx-q7jo3 1/1 Running 0 29m
|
||||
NAME READY STATUS RESTARTS AGE
|
||||
my-nginx-3800858182-h9v8d 1/1 Running 0 1m
|
||||
my-nginx-3800858182-wqafx 1/1 Running 0 1m
|
||||
```
|
||||
|
||||
Kubernetes will ensure that your application keeps running, by automatically restarting containers that fail, spreading containers across nodes, and recreating containers on new nodes when nodes fail.
|
||||
|
||||
## Exposing your application to the Internet
|
||||
|
||||
Through integration with some cloud providers (for example Google Compute Engine and AWS EC2), Kubernetes enables you to request that it provision a public IP address for your application. To do this run:
|
||||
|
||||
```shell
|
||||
$ kubectl expose rc my-nginx --port=80 --type=LoadBalancer
|
||||
service "my-nginx" exposed
|
||||
```
|
||||
|
||||
To find the public IP address assigned to your application, execute:
|
||||
|
||||
```shell
|
||||
$ kubectl get svc my-nginx
|
||||
NAME CLUSTER_IP EXTERNAL_IP PORT(S) SELECTOR AGE
|
||||
my-nginx 10.179.240.1 25.1.2.3 80/TCP run=nginx 8d
|
||||
$ kubectl get service/my-nginx
|
||||
NAME CLUSTER_IP EXTERNAL_IP PORT(S) AGE
|
||||
my-nginx 10.179.240.1 25.1.2.3 80/TCP 8s
|
||||
```
|
||||
|
||||
You may need to wait for a minute or two for the external ip address to be provisioned.
|
||||
|
|
@ -51,7 +43,7 @@ You may need to wait for a minute or two for the external ip address to be provi
|
|||
In order to access your nginx landing page, you also have to make sure that traffic from external IPs is allowed. Do this by opening a [firewall to allow traffic on port 80](/docs/user-guide/services-firewalls).
|
||||
|
||||
If you're running on AWS, Kubernetes creates an ELB for you. ELBs use host
|
||||
names, not IPs, so you will have to do `kubectl describe svc my-nginx` and look
|
||||
names, not IPs, so you will have to do `kubectl describe service/my-nginx` and look
|
||||
for the `LoadBalancer Ingress` host name. Traffic from external IPs is allowed
|
||||
automatically.
|
||||
|
||||
|
|
@ -60,12 +52,11 @@ automatically.
|
|||
To kill the application and delete its containers and public IP address, do:
|
||||
|
||||
```shell
|
||||
$ kubectl delete rc my-nginx
|
||||
replicationcontrollers/my-nginx
|
||||
$ kubectl delete svc my-nginx
|
||||
services/my-nginx
|
||||
$ kubectl delete deployment,service my-nginx
|
||||
deployment "my-nginx" deleted
|
||||
service "my-nginx" deleted
|
||||
```
|
||||
|
||||
## What's next?
|
||||
|
||||
[Learn about how to configure common container parameters, such as commands and environment variables.](/docs/user-guide/configuring-containers)
|
||||
[Learn about how to configure common container parameters, such as commands and environment variables.](/docs/user-guide/configuring-containers)
|
||||
|
|
|
|||
|
|
@ -0,0 +1,17 @@
|
|||
apiVersion: extensions/v1beta1
|
||||
kind: Deployment
|
||||
metadata:
|
||||
name: my-nginx
|
||||
spec:
|
||||
replicas: 2
|
||||
template:
|
||||
metadata:
|
||||
labels:
|
||||
run: my-nginx
|
||||
spec:
|
||||
containers:
|
||||
- name: my-nginx
|
||||
image: nginx
|
||||
ports:
|
||||
- containerPort: 80
|
||||
|
||||
|
|
@ -183,4 +183,4 @@ For more information about health checking, see [Container Probes](/docs/user-gu
|
|||
|
||||
## What's Next?
|
||||
|
||||
For a complete application see the [guestbook example](https://github.com/kubernetes/kubernetes/tree/{{page.githubbranch}}/examples/guestbook/).
|
||||
For a complete application see the [guestbook example](https://github.com/kubernetes/kubernetes/tree/{{page.githubbranch}}/examples/guestbook/).
|
||||
|
|
|
|||
Loading…
Reference in New Issue