This is an internal documentation. There is a good chance you’re looking for something else. See Disclaimer.

OpenShift Client / Basic Commands

Installation

See Setup OpenShift Client.

Login

oc login https://api.c-tocco-ocp4.tocco.ch:6443
Username: <USERNAME>
Password: <PASSWORD>

Hint

You can also specify the username using -u $USERNAME.

Changing Current Project

Just like cd $DIR can be used to change your current directory, oc project $PROJECT can be used to change the current project.

Switch to the Nice project of ${INSTALLATION}:

oc project nice-${INSTALLATION}

Show current project:

oc project

Show all projects:

oc projects

Resource Types

To get a list of all resources use oc get.

Here are the resource types you’ll need:

Type

Description

dc

Deployment configuration: A template for a pod. If you need to change a pod, change this configuration. Changes are automatically deployed.

pod

A pod is an instance of a Deployment Configuration (AKA dc).

  • Change the dc to change the configuration of a pod

route

A DNS route

oc get route
NAME      HOST/PORT          PATH      SERVICES   PORT      TERMINATION     WILDCARD
nice      abc.tocco.ch                 nice       80-tcp    edge/Redirect   None
nice      www.abc.ch                   nice       80-tcp    edge/Redirect   None

svc

Service (svc): Represents a set of pods that provide a service. Allows talking to Solr, for instance, without having to worry about in what pod it is running in or how many instances are running.

is

Image stream: This is a docker image that has been pushed to OpenShift. In the world of docker, they are called repositories.

List Resources

Use oc get TYPE to get all list of the resource of a certain type or oc get all to show all of them.

Example

$ oc get pod
NAME            READY     STATUS    RESTARTS   AGE
nice-25-kchv3   2/2       Running   0          3h
solr-2-gt5tg    1/1       Running   0          1h

Describe a Resource in Detail

Use oc describe TYPE RESOURCE to show details about a specific pod/dc/is/….

Hint

List Resources shows how to obtain the RESOURCE name.

Example

$ oc describe pod nice-25-kchv3
Name:                   nice-25-kchv3
Namespace:              nice-test212
Security Policy:        restricted
Node:                   node19.prod.zrh.appuio.ch/172.17.176.161
Start Time:             Wed, 18 Oct 2017 13:07:00 +0200
Labels:                 deployment=nice-25
                        deploymentconfig=nice
                        run=nice
…

Edit Resources

Warning

This is only suitable for debugging. Any change will be reverted on the next run of Ansible. Permanent changes should be committed to global.yml or config.yml.

Use oc edit TYPE RESOURCE to edit a specific pod/dc/is/….

Hint

List Resources shows how to obtain the RESOURCE name.

Example

  1. Open config in editor: oc edit pod nice-25-kchv3

  2. Make any changes you want to the configuration.

  3. Save changes and exit in order to trigger a deployment.

Open Shell in Pod

oc rsh -c nice PODNAME bash

-c specifies the container name, use -c nginx to enter the nginx container or oc rsh PODNAME bash to enter a Solr pod (has only one container).

Copy File from Pod

oc cp -c nice PODNAME:/path/to/file.txt ~/destination/folder/

Synchronize Folder with Pod

oc rsync -c nice PODNAME:/path/to/folder ~/destination/folder/

Manually Deploy

Deploy latest version of Nice:

oc rollout latest dc/nice

Retry Failed Deployment

Retry failed deployment of Nice:

oc rollout retry dc/nice

Open a Remote Shell in a Pod

To get a shell within a Nice pod use oc rsh -c nice POD.

Example

$ oc rsh -c nice nice-25-kchv3
nice-25-kchv3:/app $ …

Open a shell in the Nginx container using oc rsh -c nginx nice-25-kchv3 or in the Solr Pod using oc rsh solr-2-gt5tg.

Access Log Files in Nice Pod

oc exec -c nice PODNAME -- tail -n +0 var/log/nice.log |less

Scale Up/Down (Start/Stop instances)

$ oc get dc nice
NAME      REVISION   DESIRED   CURRENT   TRIGGERED BY
nice      49         2         2         config,image(nice:latest)

  • DESIRED: Number of replicas/instances configured

  • CURRENT: Number of replicas/instances currently online

Use this command to scale Nice instances:

oc scale dc/nice --replicas=${N}

This scales Nice to N replicas.

Use 0 to stop all instances. Before doing so schedule a downtime on https://monitoring.vshn.net.

Restart a Nice Instance

List pods:

$ oc get pod
NAME            READY     STATUS    RESTARTS   AGE
nice-61-6wlvx   2/2       Running   0          8d
solr-6-c6t2t    1/1       Running   0          9d

To restart the nice instance, we have to delete the pod called nice-61-6wlvx in this example.

$ oc delete pod nice-61-6wlvx
pod "nice-61-6wlvx" deleted

If we list the pods again, we see that the deleted pod is terminating and a new one has been created (nice-61-pkksw):

$ oc get pod
NAME            READY     STATUS              RESTARTS   AGE
nice-61-6wlvx   1/2       Terminating         0          8d
nice-61-pkksw   0/2       ContainerCreating   0          4s
solr-6-c6t2t    1/1       Running             0          9d

The new instance is not fully ready and, therefore, not reachable yet.

If we check again a few minutes later, the old instance has disappeared. Also, the new instance should be fully ready (and we should be able to access it again via browser):

$ oc get pod
NAME            READY     STATUS    RESTARTS   AGE
nice-61-pkksw   2/2       Running   0          7m
solr-6-c6t2t    1/1       Running   0          9d

Start PSQL (SQL Console)

See Option a) via OpenShift.

Further Commands

Look at the help output of oc to see available commands:

oc --help

The OpenShift client package also includes kubectl, a client for the Kubernetes, a core part of OpenShift:

kubectl --help