ConfigMaps

Overview

Many applications require configuration using some combination of configuration files, command line arguments, and environment variables. These configuration artifacts should be decoupled from image content in order to keep containerized applications portable.

The **ConfigMap** object provides mechanisms to inject containers with configuration data while keeping containers agnostic of OKD. A **ConfigMap** can be used to store fine-grained information like individual properties or coarse-grained information like entire configuration files or JSON blobs.

The **ConfigMap** API object holds key-value pairs of configuration data that can be consumed in pods or used to store configuration data for system components such as controllers. **ConfigMap** is similar to secrets, but designed to more conveniently support working with strings that do not contain sensitive information.

For example:

ConfigMap Object Definition

  1. kind: ConfigMap
  2. apiVersion: v1
  3. metadata:
  4.   creationTimestamp: 2016-02-18T19:14:38Z
  5.   name: example-config
  6.   namespace: default
  7. data: (1)
  8.   example.property.1: hello
  9.   example.property.2: world
  10.   example.property.file: |-
  11.     property.1=value-1
  12.     property.2=value-2
  13.     property.3=value-3
  14. binaryData:
  15.   bar: L3Jvb3QvMTAw (2)
1Contains the configuration data.
2Points to a file that contains non-UTF8 data, for example, a binary Java keystore file. Enter the file path in Base 64.

You can use the binaryData field when you create a configmap from a file.

Configuration data can be consumed in pods in a variety of ways. A **ConfigMap** can be used to:

  1. Populate the value of environment variables.

  2. Set command-line arguments in a container.

  3. Populate configuration files in a volume.

Both users and system components may store configuration data in a **ConfigMap**.

Creating ConfigMaps

You can use the following command to create a **ConfigMap** easily from directories, specific files, or literal values:

  1. $ oc create configmap <configmap_name> [options]

The following sections cover the different ways you can create a **ConfigMap**.

Creating from Directories

Consider a directory with some files that already contain the data with which you want to populate a **ConfigMap**:

  1. $ ls example-files
  2. game.properties
  3. ui.properties
  5. $ cat example-files/game.properties
  6. enemies=aliens
  7. lives=3
  8. enemies.cheat=true
  9. enemies.cheat.level=noGoodRotten
  10. secret.code.passphrase=UUDDLRLRBABAS
  11. secret.code.allowed=true
  12. secret.code.lives=30
  14. $ cat example-files/ui.properties
  15. color.good=purple
  16. color.bad=yellow
  17. allow.textmode=true
  18. how.nice.to.look=fairlyNice

You can use the following command to create a **ConfigMap** holding the content of each file in this directory:

  1. $ oc create configmap game-config \
  2.     --from-file=example-files/

When the --from-file option points to a directory, each file directly in that directory is used to populate a key in the **ConfigMap**, where the name of the key is the file name, and the value of the key is the content of the file.

For example, the above command creates the following **ConfigMap**:

  1. $ oc describe configmaps game-config
  2. Name:           game-config
  3. Namespace:      default
  4. Labels:         <none>
  5. Annotations:    <none>
  7. Data
  9. game.properties:        121 bytes
  10. ui.properties:          83 bytes

You can see the two keys in the map are created from the file names in the directory specified in the command. Because the content of those keys may be large, the output of oc describe only shows the names of the keys and their sizes.

If you want to see the values of the keys, you can oc get the object with the -o option:

  1. $ oc get configmaps game-config -o yaml
  3. apiVersion: v1
  4. data:
  5.   game.properties: |-
  6.     enemies=aliens
  7.     lives=3
  8.     enemies.cheat=true
  9.     enemies.cheat.level=noGoodRotten
  10.     secret.code.passphrase=UUDDLRLRBABAS
  11.     secret.code.allowed=true
  12.     secret.code.lives=30
  13.   ui.properties: |
  14.     color.good=purple
  15.     color.bad=yellow
  16.     allow.textmode=true
  17.     how.nice.to.look=fairlyNice
  18. kind: ConfigMap
  19. metadata:
  20.   creationTimestamp: 2016-02-18T18:34:05Z
  21.   name: game-config
  22.   namespace: default
  23.   resourceVersion: "407"-
  24.   selflink: /api/v1/namespaces/default/configmaps/game-config
  25.   uid: 30944725-d66e-11e5-8cd0-68f728db1985

Creating from Files

You can also pass the --from-file option with a specific file, and pass it multiple times to the CLI. The following yields equivalent results to the Creating from Directories example:

If you create a configmap from a file, you can include files containing non-UTF8 data will be placed in this new field without corrupting the non-UTF8 data. OKD detects binary files and transparently encodes the file as MIME. On the server, the MIME payload is decoded and stored without corrupting the data.

  1. Create the **ConfigMap** specifying a specific file:

    1. $ oc create configmap game-config-2 \
    2.     --from-file=example-files/game.properties \
    3.     --from-file=example-files/ui.properties

  2. Verify the results:

    1. $ oc get configmaps game-config-2 -o yaml
    3. apiVersion: v1
    4. data:
    5.   game.properties: |-
    6.     enemies=aliens
    7.     lives=3
    8.     enemies.cheat=true
    9.     enemies.cheat.level=noGoodRotten
    10.     secret.code.passphrase=UUDDLRLRBABAS
    11.     secret.code.allowed=true
    12.     secret.code.lives=30
    13.   ui.properties: |
    14.     color.good=purple
    15.     color.bad=yellow
    16.     allow.textmode=true
    17.     how.nice.to.look=fairlyNice
    18. kind: ConfigMap
    19. metadata:
    20.   creationTimestamp: 2016-02-18T18:52:05Z
    21.   name: game-config-2
    22.   namespace: default
    23.   resourceVersion: "516"
    24.   selflink: /api/v1/namespaces/default/configmaps/game-config-2
    25.   uid: b4952dc3-d670-11e5-8cd0-68f728db1985

You can also set the key to use for an individual file with the --from-file option by passing an expression of key=value. For example:

  1. Create the **ConfigMap** specifying a key-value pair:

    1. $ oc create configmap game-config-3 \
    2.     --from-file=game-special-key=example-files/game.properties

  2. Verify the results:

    1. $ oc get configmaps game-config-3 -o yaml
    3. apiVersion: v1
    4. data:
    5.   game-special-key: |-
    6.     enemies=aliens
    7.     lives=3
    8.     enemies.cheat=true
    9.     enemies.cheat.level=noGoodRotten
    10.     secret.code.passphrase=UUDDLRLRBABAS
    11.     secret.code.allowed=true
    12.     secret.code.lives=30
    13. kind: ConfigMap
    14. metadata:
    15.   creationTimestamp: 2016-02-18T18:54:22Z
    16.   name: game-config-3
    17.   namespace: default
    18.   resourceVersion: "530"
    19.   selflink: /api/v1/namespaces/default/configmaps/game-config-3
    20.   uid: 05f8da22-d671-11e5-8cd0-68f728db1985

Creating from Literal Values

You can also supply literal values for a **ConfigMap**. The --from-literal option takes a key=value syntax that allows literal values to be supplied directly on the command line:

  1. Create the **ConfigMap** specifying a literal value:

    1. $ oc create configmap special-config \
    2.     --from-literal=special.how=very \
    3.     --from-literal=special.type=charm

  2. Verify the results:

    1. $ oc get configmaps special-config -o yaml
    3. apiVersion: v1
    4. data:
    5.   special.how: very
    6.   special.type: charm
    7. kind: ConfigMap
    8. metadata:
    9.   creationTimestamp: 2016-02-18T19:14:38Z
    10.   name: special-config
    11.   namespace: default
    12.   resourceVersion: "651"
    13.   selflink: /api/v1/namespaces/default/configmaps/special-config
    14.   uid: dadce046-d673-11e5-8cd0-68f728db1985

Use Cases: Consuming ConfigMaps in Pods

The following sections describe some uses cases when consuming **ConfigMap** objects in pods.

Consuming in Environment Variables

**ConfigMaps** can be used to populate individual environment variables or can populate environment variables from all keys that form valid environment variable names. As an example, consider the following **ConfigMaps**:

ConfigMap with two environment variables

  1. apiVersion: v1
  2. kind: ConfigMap
  3. metadata:
  4.   name: special-config (1)
  5.   namespace: default
  6. data:
  7.   special.how: very (2)
  8.   special.type: charm (2)
1Name of the ConfigMap.
2Environment variables to inject.

ConfigMap with one environment variable

  1. apiVersion: v1
  2. kind: ConfigMap
  3. metadata:
  4.   name: env-config (1)
  5.   namespace: default
  6. data:
  7.   log_level: INFO (2)
1Name of the ConfigMap.
2Environment variable to inject.

You can consume the keys of this **ConfigMap** in a pod using **configMapKeyRef** sections:

Sample pod specification configured to inject specific environment variables

  1. apiVersion: v1
  2. kind: Pod
  3. metadata:
  4.   name: dapi-test-pod
  5. spec:
  6.   containers:
  7.     - name: test-container
  8.       image: gcr.io/google_containers/busybox
  9.       command: [ "/bin/sh", "-c", "env" ]
  10.       env: (1)
  11.         - name: SPECIAL_LEVEL_KEY
  12.           valueFrom:
  13.             configMapKeyRef:
  14.               name: special-config (2)
  15.               key: special.how (3)
  16.         - name: SPECIAL_TYPE_KEY
  17.           valueFrom:
  18.             configMapKeyRef:
  19.               name: special-config (2)
  20.               key: special.type (3)
  21.               optional: true (4)
  22.       envFrom: (5)
  23.         - configMapRef:
  24.             name: env-config (6)
  25.   restartPolicy: Never
1Stanza to pull the specified environment variables from a ConfigMap.
2Name of the ConfigMap to pull specific environment variables from.
3Environment variable to pull from the ConfigMap.
4Makes the environment variable optional. As optional, the pod will be started even if the specified ConfigMap and keys do not exist.
5Stanza to pull all environment variables from a ConfigMap.
6Name of the ConfigMap to pull all environment variables.

When this pod is run, its output will include the following lines:

  1. SPECIAL_LEVEL_KEY=very
  2. log_level=INFO

Setting Command-line Arguments

A **ConfigMap** can also be used to set the value of the command or arguments in a container. This is accomplished using the Kubernetes substitution syntax $(VAR_NAME). Consider the following **ConfigMaps**:

  1. apiVersion: v1
  2. kind: ConfigMap
  3. metadata:
  4.   name: special-config
  5.   namespace: default
  6. data:
  7.   special.how: very
  8.   special.type: charm

To inject values into the command line, you must consume the keys you want to use as environment variables, as in the Consuming in Environment Variables use case. Then you can refer to them in a container’s command using the $(VAR_NAME) syntax.

Sample pod specification configured to inject specific environment variables

  1. apiVersion: v1
  2. kind: Pod
  3. metadata:
  4.   name: dapi-test-pod
  5. spec:
  6.   containers:
  7.     - name: test-container
  8.       image: gcr.io/google_containers/busybox
  9.       command: [ "/bin/sh", "-c", "echo $(SPECIAL_LEVEL_KEY) $(SPECIAL_TYPE_KEY)" ]
  10.       env:
  11.         - name: SPECIAL_LEVEL_KEY
  12.           valueFrom:
  13.             configMapKeyRef:
  14.               name: special-config
  15.               key: special.how
  16.         - name: SPECIAL_TYPE_KEY
  17.           valueFrom:
  18.             configMapKeyRef:
  19.               name: special-config
  20.               key: special.type
  21.   restartPolicy: Never

When this pod is run, the output from the test-container container will be:

  1. very charm

Consuming in Volumes

A **ConfigMap** can also be consumed in volumes. Returning again to the following example **ConfigMap**:

  1. apiVersion: v1
  2. kind: ConfigMap
  3. metadata:
  4.   name: special-config
  5.   namespace: default
  6. data:
  7.   special.how: very
  8.   special.type: charm

You have a couple different options for consuming this **ConfigMap** in a volume. The most basic way is to populate the volume with files where the key is the file name and the content of the file is the value of the key:

  1. apiVersion: v1
  2. kind: Pod
  3. metadata:
  4.   name: dapi-test-pod
  5. spec:
  6.   containers:
  7.     - name: test-container
  8.       image: gcr.io/google_containers/busybox
  9.       command: [ "/bin/sh", "cat", "/etc/config/special.how" ]
  10.       volumeMounts:
  11.       - name: config-volume
  12.         mountPath: /etc/config
  13.   volumes:
  14.     - name: config-volume
  15.       configMap:
  16.         name: special-config
  17.   restartPolicy: Never

When this pod is run, the output will be:

  1. very

You can also control the paths within the volume where **ConfigMap** keys are projected:

  1. apiVersion: v1
  2. kind: Pod
  3. metadata:
  4.   name: dapi-test-pod
  5. spec:
  6.   containers:
  7.     - name: test-container
  8.       image: gcr.io/google_containers/busybox
  9.       command: [ "/bin/sh", "cat", "/etc/config/path/to/special-key" ]
  10.       volumeMounts:
  11.       - name: config-volume
  12.         mountPath: /etc/config
  13.   volumes:
  14.     - name: config-volume
  15.       configMap:
  16.         name: special-config
  17.         items:
  18.         - key: special.how
  19.           path: path/to/special-key
  20.   restartPolicy: Never

When this pod is run, the output will be:

  1. very

Example: Configuring Redis

For a real-world example, you can configure Redis using a **ConfigMap**. To inject Redis with the recommended configuration for using Redis as a cache, the Redis configuration file should contain the following:

  1. maxmemory 2mb
  2. maxmemory-policy allkeys-lru

If your configuration file is located at example-files/redis/redis-config, create a **ConfigMap** with it:

  1. Create the **ConfigMap** specifying the configuration file:

    1. $ oc create configmap example-redis-config \
    2.     --from-file=example-files/redis/redis-config

  2. Verify the results:

    1. $ oc get configmap example-redis-config -o yaml
    3. apiVersion: v1
    4. data:
    5.   redis-config: |
    6.     maxmemory 2mb
    7.     maxmemory-policy allkeys-lru
    8. kind: ConfigMap
    9. metadata:
    10.   creationTimestamp: 2016-04-06T05:53:07Z
    11.   name: example-redis-config
    12.   namespace: default
    13.   resourceVersion: "2985"
    14.   selflink: /api/v1/namespaces/default/configmaps/example-redis-config
    15.   uid: d65739c1-fbbb-11e5-8a72-68f728db1985

Now, create a pod that uses this **ConfigMap**:

  1. Create a pod definition like the following and save it to a file, for example redis-pod.yaml:

    1. apiVersion: v1
    2. kind: Pod
    3. metadata:
    4.   name: redis
    5. spec:
    6.   containers:
    7.   - name: redis
    8.     image: kubernetes/redis:v1
    9.     env:
    10.     - name: MASTER
    11.       value: "true"
    12.     ports:
    13.     - containerPort: 6379
    14.     resources:
    15.       limits:
    16.         cpu: "0.1"
    17.     volumeMounts:
    18.     - mountPath: /redis-master-data
    19.       name: data
    20.     - mountPath: /redis-master
    21.       name: config
    22.   volumes:
    23.     - name: data
    24.       emptyDir: {}
    25.     - name: config
    26.       configMap:
    27.         name: example-redis-config
    28.         items:
    29.         - key: redis-config
    30.           path: redis.conf

  2. Create the pod:

    1. $ oc create -f redis-pod.yaml

The newly-created pod has a **ConfigMap** volume that places the redis-config key of the example-redis-config **ConfigMap** into a file called redis.conf. This volume is mounted into the /redis-master directory in the Redis container, placing our configuration file at /redis-master/redis.conf, which is where the image looks for the Redis configuration file for the master.

If you oc exec into this pod and run the redis-cli tool, you can check that the configuration was applied correctly:

  1. $ oc exec -it redis redis-cli
  2. 127.0.0.1:6379> CONFIG GET maxmemory
  3. 1) "maxmemory"
  4. 2) "2097152"
  5. 127.0.0.1:6379> CONFIG GET maxmemory-policy
  6. 1) "maxmemory-policy"
  7. 2) "allkeys-lru"

Restrictions

A **ConfigMap** must be created before they are consumed in pods. Controllers can be written to tolerate missing configuration data; consult individual components configured via **ConfigMap** on a case-by-case basis.

**ConfigMap** objects reside in a project. They can only be referenced by pods in the same project.

The Kubelet only supports use of a **ConfigMap** for pods it gets from the API server. This includes any pods created using the CLI, or indirectly from a replication controller. It does not include pods created using the OKD node’s --manifest-url flag, its --config flag, or its REST API (these are not common ways to create pods).