Update a Deployment

Change the image version, replica count, or environment variables of a running Deployment. All changes are applied as a rolling update with no service downtime.

---

Step 1. Open the action menu

1. In the left menu, select Workloads → Deployments.
2. Click the action button (⋮) at the right end of the target Deployment row.

The action menu includes:

Menu item	Description
Execute shell	Shell (exec) access to a Pod container
Pause orchestration	Pause the rollout
Redeploy	Redeploy with the current spec (restarts Pods)
Rollback	Revert to the previous revision
Edit config	Open the config edit form
Edit YAML	Open the YAML editor
Download YAML	Download the current spec as YAML
Delete	Delete the Deployment

---

Edit config (form)

Edit config opens a full edit form — the same layout as the create form — with three tabs (Deployment / Pod / Container) and a section for every part of the spec. Click Edit config from the action menu, change values across the tabs, then click Save in the Summary panel. To edit the raw manifest instead, use Edit YAML (a separate screen).

Edit Config: Deployment

Deployment-level settings — replica count and selector labels, plus the scaling/upgrade policy at the bottom of the tab.

#	Field	Description
2-1	Replicas	Number of Pod replicas to run (integer ≥ 0), in Basic information.
2-2	Selector labels	Pod-selector Match Labels (immutable after creation).
2-3	Save	Save changes; applied as a rolling update.

#	Field	Description	Zero-downtime
3-1	Update policy	Rolling Update (zero-downtime) or On Delete.	Rolling Update
3-2	Max surge	Max extra Pods created during an update (count or %).	1
3-3	Max unavailable	Max existing Pods unavailable during an update (count or %).	0
3-4	Minimum ready	Min seconds a new Pod must stay Ready to count as available.	—
3-5	Revision history limit	Number of past revisions kept for rollback.	—
3-6	Progress deadline	Max seconds allowed before the rollout is marked failed.	—

Zero-downtime

With Update policy = Rolling Update, Max unavailable = 0, and Max surge = 1, an old Pod is terminated only after the new Pod is Ready, so there is no service interruption.

Edit Config: Pod

Pod-template settings shared by all containers.

#	Section	Description
4-1	Termination	Pod termination grace period.
4-2	Networking	Network mode, DNS policy, hostname, etc.
5-1	Node scheduling	Node placement rules (nodeSelector / affinity).
5-2	Pod scheduling	Priority, tolerations, and scheduling options.
6-1	Resources	Pod-level resource settings.
6-2	Security context	Pod security context (runAsUser, fsGroup, …).
6-3	Storage	Pod-level volumes / storage.

Edit Config: Container

Select the container-name tab to edit that container's spec.

#	Section	Description
7-1	Image	Container image, Pull Policy, Pull Secrets. Changing the image triggers a rolling update.
7-2	Networking	Service and ports the container exposes.
8-1	Command	Override container Command / Args.
8-2	Environment variables	Add / edit / remove env vars.
8-3	Service Account	Service Account used by the container.
9-1	Lifecycle hooks	postStart / preStop hooks.
9-2	Resources	CPU / Memory reservations and limits.
10-1	Security context	Container security context (privileges, capabilities).
10-2	Health check	Liveness / Readiness / Startup probes.
10-3	Storage	Volume mounts and container storage.

---

Rollback to a previous version

1. Click Rollback from the action menu.
2. Confirm in the dialog by clicking Confirm.

The previous ReplicaSet is restored immediately.

---

Redeploy (force restart)

Restart all Pods without changing the spec (e.g., to refresh cached images).

1. Click Redeploy from the action menu.
2. Confirm in the dialog.

WARNING

Redeploy replaces all Pods. When used with imagePullPolicy: Always, it forces a fresh image pull.

---

Next steps

• View Pod logs →
• Manage Pods →