Why Feature flags in Kubernetes? - Purpose & Use Cases

The Big Idea

What if you could turn new features on or off instantly without touching your code or restarting anything?

The Scenario

Imagine you want to test a new feature in your app running on Kubernetes. You have to change the code, rebuild the container, update the deployment, and hope nothing breaks. If you want to turn the feature off, you repeat the whole process.

The Problem

This manual way is slow and risky. Every change means downtime or bugs. You can't quickly switch features on or off for different users or environments. It feels like juggling many balls and dropping some.

The Solution

Feature flags in Kubernetes let you control features without changing code or redeploying. You toggle features on or off by changing simple settings. This makes testing, rolling out, or rolling back features fast and safe.

Before vs After

✗ Before

kubectl set image deployment/myapp myapp=myapp:v2
kubectl rollout restart deployment/myapp

✓ After

kubectl patch configmap feature-flags -p '{"data":{"newFeature":"enabled"}}'

What It Enables

You can safely test and release features anytime, anywhere, with zero downtime and full control.

Real Life Example

A team wants to try a new payment option only for 10% of users. Using feature flags in Kubernetes, they enable it just for that group without affecting others or redeploying.

Key Takeaways

Manual feature changes cause delays and risks.

Feature flags let you toggle features instantly without redeploying.

This improves safety, speed, and flexibility in managing app features.

Practice

1. What is the main purpose of feature flags in Kubernetes?

easy

A. To manage Kubernetes cluster nodes

B. To monitor cluster health

C. To schedule pods across nodes

D. To enable or disable application features without changing code

Solution

1. Step 1: Understand feature flags concept

   Feature flags allow toggling features on or off without modifying the application code.

2. Step 2: Relate to Kubernetes usage

   In Kubernetes, feature flags help control app behavior dynamically, often via ConfigMaps or environment variables.

3. Final Answer:

   To enable or disable application features without changing code -> Option D

4. Quick Check:

   Feature flags = toggle features without code change [OK]

Hint: Feature flags toggle features without code edits [OK]

Common Mistakes:

• Confusing feature flags with cluster management
• Thinking feature flags manage pods or nodes
• Mixing feature flags with monitoring tools

2. Which Kubernetes resource is commonly used to store feature flags for an application?

easy

A. Service

B. Pod

C. ConfigMap

D. Ingress

Solution

1. Step 1: Identify resource types

   Pods run containers, Services expose them, Ingress manages external access, ConfigMaps store configuration data.

2. Step 2: Match feature flags storage

   Feature flags are configuration data, so ConfigMaps are the right resource to store them.

3. Final Answer:

   ConfigMap -> Option C

4. Quick Check:

   Feature flags stored in ConfigMap [OK]

Hint: ConfigMaps hold config data like feature flags [OK]

Common Mistakes:

• Choosing Pod instead of ConfigMap
• Confusing Service or Ingress with config storage
• Thinking feature flags are stored in Secrets

3. Given this ConfigMap YAML snippet for feature flags:

apiVersion: v1
kind: ConfigMap
metadata:
  name: feature-flags
data:
  FEATURE_X_ENABLED: "true"
  FEATURE_Y_ENABLED: "false"

What will be the value of FEATURE_Y_ENABLED when accessed as an environment variable in a pod?

medium

A. false

B. true

C. null

D. undefined

Solution

1. Step 1: Read ConfigMap data values

   The ConfigMap sets FEATURE_Y_ENABLED to the string "false" explicitly.

2. Step 2: Understand environment variable mapping

   When injected as env vars, values are strings exactly as in ConfigMap, so FEATURE_Y_ENABLED will be "false".

3. Final Answer:

   false -> Option A

4. Quick Check:

   ConfigMap value "false" = env var "false" [OK]

Hint: Env vars get exact string values from ConfigMap [OK]

Common Mistakes:

• Assuming boolean false instead of string "false"
• Thinking missing keys return null or undefined
• Confusing string values with boolean types

4. You have this environment variable setup in a pod spec:

- name: FEATURE_Z_ENABLED
  valueFrom:
    configMapKeyRef:
      name: feature-flags
      key: FEATURE_Z_ENABLED

If the ConfigMap 'feature-flags' does not have the key FEATURE_Z_ENABLED, what will happen when the pod starts?

medium

A. Pod will fail to start with an error

B. Pod will start with FEATURE_Z_ENABLED set to an empty string

C. Pod will start with FEATURE_Z_ENABLED set to null

D. Pod will ignore the environment variable

Solution

1. Step 1: Understand configMapKeyRef behavior

   If the specified key is missing in the ConfigMap, Kubernetes treats it as an error.

2. Step 2: Effect on pod startup

   The pod will fail to start because the environment variable cannot be resolved from the ConfigMap key.

3. Final Answer:

   Pod will fail to start with an error -> Option A

4. Quick Check:

   Missing ConfigMap key causes pod start failure [OK]

Hint: Missing ConfigMap key breaks pod start [OK]

Common Mistakes:

• Assuming empty string or null is set silently
• Thinking pod ignores missing keys
• Confusing with optional environment variables

5. You want to enable a new feature only for 10% of users using feature flags in Kubernetes. Which approach best supports this scenario?

hard

A. Use a ConfigMap with a boolean flag set to true or false

B. Store a percentage value in ConfigMap and let the app decide feature enablement per user

C. Use a Secret to store the feature flag and update it daily

D. Deploy two versions of the app and route 10% traffic to the new version

Solution

1. Step 1: Understand percentage-based feature flags

   To enable a feature for a subset of users, the flag must support partial enablement, not just true/false.

2. Step 2: Evaluate options

   Store a percentage value in ConfigMap and let the app decide feature enablement per user stores a percentage in ConfigMap; the app reads it and enables the feature for that percent of users dynamically.

3. Final Answer:

   Store a percentage value in ConfigMap and let the app decide feature enablement per user -> Option B

4. Quick Check:

   Percentage flags need app logic with ConfigMap value [OK]

Hint: Use percentage value in ConfigMap for partial rollout [OK]

Common Mistakes:

• Using boolean flags for partial user enablement
• Relying on Secrets for feature flags
• Thinking traffic routing replaces feature flags