# Horizontal Manual Scaling for a Deployment LLMS index: [llms.txt](/llms.txt) --- This page shows how to manually scale a Deployment horizontally, by changing its replica count. Manual scaling lets you directly control the number of running Pods for predictable load changes or cost management. This is different from _vertical scaling_: leaving the replica count the same, but adjusting the amount of resources available to each Pod. ## Objectives - Scaling up a Deployment to handle more traffic. - Scaling down a Deployment to conserve resources. - Scaling a Deployment to zero to suspend a workload. - Understanding when to use manual scaling versus a HorizontalPodAutoscaler. ## Before you begin

You need to have a Kubernetes cluster, and the kubectl command-line tool must be configured to communicate with your cluster. It is recommended to run this tutorial on a cluster with at least two nodes that are not acting as control plane hosts. If you do not already have a cluster, you can create one by using minikube or you can use one of these Kubernetes playgrounds:

You need an existing Deployment. If you do not have one, and you just want to practice, you can create the nginx Deployment from [Run a Stateless Application Using a Deployment](/docs/tasks/run-application/run-stateless-application-deployment/): ```shell kubectl apply -f https://k8s.io/examples/application/deployment.yaml ``` Verify the Deployment runs two Pods: ```shell kubectl get deployment nginx-deployment ``` The output is similar to: ``` NAME READY UP-TO-DATE AVAILABLE AGE nginx-deployment 2/2 2 2 10s ``` ## Scaling up a Deployment There are several different ways you can change the replica count for an existing Deployment. ### Scaling up using `kubectl scale` Use `kubectl scale` to set the replica count: ```shell kubectl scale deployment/nginx-deployment --replicas=4 ``` The output is similar to: ``` deployment.apps/nginx-deployment scaled ``` Verify that the Deployment has four Pods: ```shell kubectl get deployment nginx-deployment ``` The output is similar to: ``` NAME READY UP-TO-DATE AVAILABLE AGE nginx-deployment 4/4 4 4 1m ``` ### Declarative scaling using `kubectl apply` Instead of running an imperative command, you can update the manifest file and apply it. This approach fits well with version-controlled configuration workflows. Save the current Deployment configuration to a local file: ```shell kubectl get deployment nginx-deployment -o yaml > /tmp/nginx-deployment.yaml ``` Edit `/tmp/nginx-deployment.yaml` and change `.spec.replicas` to `4`. Before applying, compare your local changes against the cluster state: ```shell kubectl diff -f /tmp/nginx-deployment.yaml ``` Apply the edited manifest: ```shell kubectl apply -f /tmp/nginx-deployment.yaml ``` ## Scaling down a Deployment To reduce the number of Pods, set `--replicas` to a lower value: ```shell kubectl scale deployment/nginx-deployment --replicas=2 ``` Kubernetes gracefully terminates the excess Pods, respecting each Pod's `terminationGracePeriodSeconds` setting. Verify that the Deployment has two Pods: ```shell kubectl get pods -l app=nginx ``` The output is similar to: ``` NAME READY STATUS RESTARTS AGE nginx-deployment-66b6c48dd5-7gl6h 1/1 Running 0 2m nginx-deployment-66b6c48dd5-v8mkd 1/1 Running 0 2m ``` ## Scaling to zero You can scale a Deployment to zero to temporarily suspend the workload without deleting the Deployment itself: ```shell kubectl scale deployment/nginx-deployment --replicas=0 ``` Verify that no Pods are running: ```shell kubectl get deployment nginx-deployment ``` The output is similar to: ``` NAME READY UP-TO-DATE AVAILABLE AGE nginx-deployment 0/0 0 0 5m ```

Note:

Scaling to zero removes all Pods but preserves the Deployment and its ReplicaSet. Scale back up at any time by setting --replicas to a positive number.
Common use cases for scaling to zero include: - Temporarily suspending a workload to save resources - Debugging or maintenance windows - Cost control in development or staging environments ## Other ways to change the replica count In addition to `kubectl scale`, you can change `.spec.replicas` with `kubectl edit` or `kubectl patch`. ### Scale using `kubectl edit` ```shell kubectl edit deployment nginx-deployment ``` Change the `.spec.replicas` field in the editor, then save and exit. ### Scale using `kubectl patch` You can update `.spec.replicas` with a strategic merge patch: ```shell kubectl patch deployment nginx-deployment -p '{"spec":{"replicas":4}}' ``` For scripting, use a JSON patch with a prerequisite test. The following command sets the replica count to 4, but only if the current count is 2: ```shell kubectl patch deployment nginx-deployment --type=json -p='[ {"op": "test", "path": "/spec/replicas", "value": 2}, {"op": "replace", "path": "/spec/replicas", "value": 4} ]' ``` The `test` operation causes the patch to fail if the current value does not match, which prevents unintended changes when multiple people or scripts modify the same Deployment. ## When to use manual versus automatic scaling | Aspect | Manual scaling | Automatic scaling (HPA) | |--------|---------------|------------------------| | Best for | Predictable, scheduled, or one-off load changes | Variable or unpredictable demand | | How it works | You set `.spec.replicas` directly | HPA adjusts replicas based on observed metrics | | Response time | Immediate when you run the command | Reacts to metrics with a short delay | | Metrics awareness | None — you decide the replica count | Monitors CPU, memory, or custom metrics | | Maintenance | Requires manual intervention to adjust | Runs autonomously after configuration |

Caution:

If a HorizontalPodAutoscaler manages a Deployment, do not set replicas manually. The HPA continuously reconciles the replica count and overrides any manual changes.
## Cleaning up Delete the Deployment: ```shell kubectl delete deployment nginx-deployment ``` ## What's next - Learn more about [Deployments](/docs/concepts/workloads/controllers/deployment/). - Walk through [Horizontal Pod Autoscaling](/docs/tasks/run-application/horizontal-pod-autoscale-walkthrough/). - Learn how to [scale a StatefulSet](/docs/tasks/run-application/scale-stateful-set/). - Read about [managing resources](/docs/concepts/cluster-administration/manage-deployment/).