Ntfy on GKE Autopilot

ntfy is an open-source, Apache 2.0-licensed pub/sub push-notification server written in Go. Applications publish messages over a simple REST/HTTP API and clients receive them instantly over WebSocket or Server-Sent-Events (SSE) streams — no external database required. This module deploys ntfy on GKE Autopilot on top of the App_GKE foundation, which provisions and manages the shared Google Cloud and Kubernetes infrastructure.

This guide focuses on the cloud services ntfy uses and how to explore and operate them from the Google Cloud Console and the command line. For the mechanics common to every GKE application — Workload Identity, ingress, autoscaling, CI/CD, Cloud Armor, IAP, Binary Authorization, VPC Service Controls, backups, and the deployment lifecycle — refer to the App_GKE foundation guide rather than repeating them here.

---

What Ntfy costs on RAD, and how that compares

Ntfy on RAD's GKE Autopilot module costs about US$40–113 a month, with the module fee charged once in RAD credits. GKE Autopilot suits Ntfy that must stay up, scale across pods or run beside other Kubernetes workloads; for the lowest cost, the Cloud Run module runs the same Ntfy for about US$51 a month. If you only need it occasionally — a demo, a seasonal project, something you're evaluating — a RAD-managed project can be deleted and restored within 30 days for a handful of credits (under US$1) instead of paying for it to sit idle. See Pause it for free, below. Figures are as at 8 October 2026; sources are listed at the end of this section.

What you pay on RAD

RAD deploys Ntfy, which is open source, so there is no licence fee per user. You pay for three things:

	In a Google Cloud project you own	In a project RAD manages for you
Module fee, once per deployment	40 credits (US$4 at the top-up price)	36 credits (10% lower)
Build time	About 3-6 credits per build	The same
Google Cloud running cost	Billed by Google to your own billing account, about US$40 a month for the default configuration (table below) — more if this isn't your only Autopilot cluster	Metered hourly in credits; RAD publishes 63 credits a day, about 1,890 a month (about US$189 at the top-up price, US$151 at the Scale plan's rate)
Who owns what	You own the project and its billing; RAD deploys and updates it	RAD owns the project, with guardrails, quotas and budget alerts; you get console access, and a minimum purchased balance is held in reserve

10 credits cost US$1 on a one-off top-up, and less on a monthly plan. In a project RAD manages, the database, file server and network are shared by every application in the project, so a second application does not add a second set of them.

Default running cost in your own project (us-central1, Google list prices):

Resource (module default)	Per month
GKE Autopilot pod, 1x vCPU / 0.5 GiB	US$34.29
Cloud NAT and networking	US$5
Cloud Storage (add-ons, backups)	US$1
GKE cluster management fee	shared across every GKE app in the project; $0 if this is your only cluster, else ~73
Total	about US$40 (about US$113 if this isn't your only Autopilot cluster)

How it compares

Option	Price	Who runs it
ntfy.sh, Supporter	$6/month ($5 billed annually)	The ntfy maintainer's own hosted instance. 2,500 messages/day, 3 reserved topics
ntfy.sh, Pro	$12/month ($10 billed annually)	20,000 messages/day, 10 reserved topics
ntfy.sh, Business	$25/month ($20 billed annually)	50,000 messages/day, 50 reserved topics, phone calls
RAD, Cloud Run, your own project	About US$51, plus US$4 once per deployment	RAD's automation, in your project
RAD, Cloud Run, a RAD-managed project	About 750 credits (US$60–75)	RAD
RAD, GKE Autopilot	See the GKE guide	RAD

ntfy's own maintainer sells exactly this — a hosted copy of the same open-source server — so it is the fairest comparison available. The free, self-hosted version has no message cap at all, which is the usual reason to self-host rather than subscribe: a busy integration can outgrow even the Business plan's 50,000 messages/day, while RAD's own project has no such ceiling.

Pause it for free: delete a RAD-managed project, restore it when you need it

If Ntfy runs in a project RAD manages for you, you have a second option that goes well beyond scaling to zero: delete the whole project, and restore it within 30 days for close to nothing. This suits Ntfy you only need occasionally — a demo, an evaluation, a seasonal project — far better than running it continuously.

• How it works. Deleting a RAD-managed project unlinks its billing first, then asks Google to delete the project. Google does not remove the project immediately: it keeps it, recoverable, for 30 days. Because billing is already unlinked, nothing is charged while it waits. Unlike deleting one module, this does not tear down the cluster or the compute resource one by one — the whole project, and everything in it, simply stops.
• Restoring costs a few credits, not a rebuild. Within 30 days, the project's owner can restore it. RAD asks Google to undelete the project and reattaches its billing account, then asks you to run Update on each deployment to confirm everything came back. Because nothing was individually destroyed, that Update finds the same resources already there — it is a check, not a rebuild, and an Update never charges the module fee again. That costs roughly a handful of credits (under US$1) in total for a typical 2-3-deployment chain.
• So a month of occasional use can cost a few dollars, not US$40. Deploy Ntfy, use it for a while, delete the project. Restore it next time you want it, confirm with Update, and delete it again when you're done. You pay only for the module fee once, the builds, and whatever time Ntfy was actually live.
• What this needs. You must own the project (not a bring-your-own one RAD only manages billing for), and you restore it yourself within the 30 days — after that, Google deletes it for good. Restoring is admitted like creating a new project: your purchased credit balance must still clear the tier's floor (100 credits for the sandbox tier most study and demo use fits). Google says most services are fully working again within 36 hours of a restore.
• One real gap: nightly backups don't survive. Backups are written to a bucket inside the project, and that bucket is not protected by Cloud Storage's soft-delete, so it is very likely gone as soon as you delete the project — even though the project itself is recoverable for 30 days. If you've customised Ntfy and want to keep that work, copy a backup out (to Google Drive, or a bucket outside the project) before you delete. For a default installation with nothing irreplaceable in it, this does not matter.

Pay only while you use it, the other way: delete and redeploy

The option above only applies to a RAD-managed project; in your own project, or once the 30-day window has passed, the way to stop paying is to delete the deployment and deploy it again when you need it.

• What a redeploy costs. The module fee again plus the builds: about 45 credits (US$4.50) in your own project, or 41 credits in a RAD-managed one. Deleting saves money once Ntfy would otherwise sit unused for about 3 days or more in your own project (about US$1.34 a day), or about 1 days or more in a RAD-managed one (63 credits a day).
• Most of the running cost is usually the database and any shared file/cache VM. They stop only when nothing else in the project uses them, so deleting Ntfy while something else shares the project saves only this app's own compute part.
• Keep your data first. Nightly backups go to a bucket inside the deployment and are deleted with it, so copy the latest backup out before deleting if you want to keep it.

Lab sessions and Managed Environments

• Lab sessions, for training. A trainer runs a session for a class. Each participant gets Ntfy in their own Google Cloud project for 15 minutes to 24 hours, within an allowance the trainer sets. Either the trainer funds every place, or each participant pays for their own. Everything is deleted when the session ends and unused credits go back to the trainer.
• Managed Environments, for consultancies. A partner runs Ntfy for a client from a ring-fenced wallet it funds, and settles with the client directly. If the wallet runs low, billing pauses and the data is kept, so nobody receives an unexpected charge. At the end the partner hands the project over and the deployments become the client's own.
• Credits can be bought in more than 20 currencies, including XAF, XOF, NGN, GHS, KES and ZAR, by card, bank transfer or mobile money.

Sources (8 October 2026): Google Cloud list prices from the Cloud Billing Catalog API; Delete and restore projects for the 30-day recovery window and what it says about Cloud Storage objects without soft delete; RAD fees and the daily-credit estimates from radmodules.dev/pricing; ntfy.sh. Prices change; check each source before relying on a figure.

1. Overview

ntfy runs as a single Go web workload. The deployment wires together a deliberately small set of Google Cloud services — ntfy has no database, cache, or object-storage dependency of its own:

Capability	Google Cloud service	Notes
Compute	GKE Autopilot	Single Go Deployment, 1 vCPU / 512 MiB by default
Database	None	database_type = "NONE"; message cache is a local SQLite file, no Cloud SQL provisioned
Persistence	Ephemeral disk (default), NFS, or block PVC (StatefulSet)	SQLite cache at /var/cache/ntfy/cache.db; NFS or PVC for durable history
Object storage	None	ntfy stores nothing in Cloud Storage
Cache / queue	None	No Redis; ntfy uses an in-process message bus
Secrets	Secret Manager	No auto-generated secrets; only user-supplied secret_environment_variables
Ingress	Cloud Load Balancing	External LoadBalancer Service; optional custom domain + managed certificate

Sensible defaults worth knowing up front:

• No database is provisioned. database_type = "NONE" — ntfy keeps its message cache in a local SQLite file. The database-related variables exist for completeness but are inert unless you deliberately opt in to an external database.
• Stateless Deployment by default. workload_type = "Deployment" with an ephemeral SQLite cache at /var/cache/ntfy/cache.db. Message history is lost when a pod restarts. For durability, either enable NFS (enable_nfs = true) or switch to a StatefulSet block PVC (stateful_pvc_enabled = true with stateful_pvc_mount_path = "/var/cache/ntfy").
• Single replica by default (min_instance_count = 1, max_instance_count = 1). A subscriber's stream is anchored to the pod that holds it and there is no shared message bus, so scaling out is not the default. Keep max at 1 unless you place a shared cache/broker behind ntfy.
• Exposed via a LoadBalancer Service (service_type = "LoadBalancer", reserve_static_ip = true, enable_custom_domain = true), so publishers and subscribers can reach it from outside the cluster.
• The health endpoint is /v1/health, which returns {"healthy":true} with HTTP 200 as soon as the server binds its port.
• The GKE variant runs on its own tenant namespace. Ntfy_GKE appends -gke to tenant_id, so it can run alongside Ntfy_CloudRun on the same tenant without a naming collision.
• Access control is a post-deploy step. ntfy ships with open access; configure users and topic ACLs afterwards via its CLI or NTFY_AUTH_* environment variables.

---

2. Google Cloud Services & How to Explore Them

All commands assume you have run gcloud container clusters get-credentials <cluster> --region <region> --project <project> and that PROJECT, REGION, and NAMESPACE are set. The namespace and other identifiers are reported in the deployment Outputs.

A. GKE Autopilot — the ntfy workload

ntfy pods are scheduled on Autopilot, which bills for the CPU/memory the pods actually request. By default the workload is a stateless Deployment with a single replica; switching to a StatefulSet provisions a per-pod block PVC for durable message history.

• Console: Kubernetes Engine → Workloads → select the ntfy workload to see pods and events. Kubernetes Engine → Services & Ingress shows the external IP.
• CLI:

  kubectl get pods,svc -n "$NAMESPACE"
  kubectl logs -n "$NAMESPACE" deploy/<service-name> --tail=100
  kubectl get statefulset -n "$NAMESPACE"          # when stateful_pvc_enabled = true

See App_GKE for how Autopilot, scaling, and the workload type (Deployment vs StatefulSet) are managed.

B. Persistence — the SQLite message cache

ntfy has no Cloud SQL instance. Its message cache is a local SQLite file at NTFY_CACHE_FILE (/var/cache/ntfy/cache.db), created by the entrypoint on boot. With the default stateless Deployment the cache is ephemeral — message history does not survive a pod restart. Two ways to make it durable:

• NFS (Filestore): enable_nfs = true, and point the cache directory at the mount.

• StatefulSet block PVC: stateful_pvc_enabled = true with stateful_pvc_mount_path = "/var/cache/ntfy" — a per-pod persistent disk holds the cache database.

• Console: Filestore → Instances (NFS); Kubernetes Engine → Storage → PersistentVolumeClaims (block PVC).

• CLI:

  kubectl get pvc -n "$NAMESPACE"                                   # StatefulSet PVCs
  kubectl exec -n "$NAMESPACE" <pod> -- ls -l /var/cache/ntfy       # cache location

See App_GKE for the NFS and StatefulSet PVC models.

C. Secret Manager

ntfy generates no secrets at deploy time — there is no database password or encryption key to manage. Secret Manager is used only if you supply your own via secret_environment_variables (for example an NTFY_AUTH_* value or an upstream push credential).

• Console: Security → Secret Manager.
• CLI:

  gcloud secrets list --project "$PROJECT"
  gcloud secrets versions access latest --secret=<secret-name> --project "$PROJECT"

See App_GKE for the Secret Store CSI integration and rotation.

D. Networking & ingress

By default the workload is exposed through an external Cloud Load Balancing IP via a LoadBalancer Service. A custom domain with a Google-managed certificate can be enabled, and a static IP is reserved by default so the address survives redeploys. If clients use HTTP/2 streaming, set container_protocol = "h2c".

• Console: Network services → Load balancing; VPC network → IP addresses.
• CLI:

  kubectl get ingress,svc -n "$NAMESPACE"
  gcloud compute addresses list --project "$PROJECT"

See App_GKE for custom domains, Cloud CDN, and static IP details.

E. Cloud Logging & Monitoring

Pod stdout/stderr flow to Cloud Logging; GKE metrics flow to Cloud Monitoring. Optional uptime checks and alert policies are available. ntfy logs its listen address and resolved cache path on startup.

• Console: Logging → Logs Explorer; Monitoring → Dashboards / Alerting.
• CLI:

  gcloud logging read 'resource.type="k8s_container" AND resource.labels.namespace_name="'"$NAMESPACE"'"' \
    --project "$PROJECT" --limit 50

---

3. Ntfy Application Behaviour

• No first-deploy database setup. ntfy has no external database and no migration step. The entrypoint prepares the SQLite cache directory and immediately execs ntfy serve. There is no db-init job by default.
• Persistence depends on the workload type. A stateless Deployment uses an ephemeral cache; a StatefulSet with a block PVC (or an NFS mount) makes message history durable across pod restarts.
• imagePullPolicy = Always for the custom image. App_GKE sets this for custom-built/mirrored images so a rebuild-and-redeploy under an unchanged tag pulls the new layers rather than serving a stale node cache.
• Health path. Startup and liveness probes target /v1/health, which returns {"healthy":true} and HTTP 200 as soon as the server binds port 80. Verify from inside the cluster:

  kubectl run curl --rm -it --image=curlimages/curl -n "$NAMESPACE" -- \
    curl -s http://<service-name>.$NAMESPACE.svc.cluster.local/v1/health   # {"healthy":true}

• Publish / subscribe smoke test (against the external IP):

  EXTERNAL_IP=$(kubectl get svc <service-name> -n "$NAMESPACE" \
    -o jsonpath='{.status.loadBalancer.ingress[0].ip}')
  curl -d "hello from ntfy" "http://$EXTERNAL_IP/mytopic"    # publish

• Access is open until you lock it down. By default any client can publish to and subscribe from any topic. Configure users and per-topic ACLs post-deploy via ntfy's CLI (ntfy user add, ntfy access) or the NTFY_AUTH_* environment variables.
• Public base URL for attachments / web push. If you use attachments or browser web-push, set NTFY_BASE_URL (via environment_variables) to the external URL so generated links resolve correctly.

---

4. Configuration Variables

Variables are grouped exactly as they appear on the deployment platform. Only settings specific to or notable for ntfy are listed; every other input is inherited from App_GKE with its standard behaviour and defaults.

Group 1 — Project & Identity

Variable	Default	Description
project_id	(required)	Target Google Cloud project.
region	us-central1	Region for the workload and regional resources.

Group 2 / 3 — Deployment Environment & Application Identity

Variable	Default	Description
tenant_id	demo	Short suffix; Ntfy_GKE appends -gke internally so it can coexist with the Cloud Run variant.
application_name	ntfy	Base name for resources. Do not change after first deploy.
application_version	latest	Image version tag; latest maps to a pinned v2.11.0 base. Pin v2.x.y in production.

Group 4 — Runtime & Scaling

Variable	Default	Description
deploy_application	true	Set false to provision infrastructure only.
container_image_source	custom	custom builds the wrapper image via Cloud Build; prebuilt deploys an image directly.
min_instance_count	1	Minimum replicas (GKE has no scale-to-zero).
max_instance_count	1	Keep at 1 — streams are pod-local with no shared broker.
container_resources	{ cpu_limit="1000m", memory_limit="512Mi" }	Per-pod CPU/memory.
container_port	80	ntfy listens on port 80.
container_protocol	http1	Set h2c for end-to-end HTTP/2 streaming.
workload_type	Deployment	Stateless default; StatefulSet for a durable per-pod PVC.
enable_cloudsql_volume	false	Off — ntfy has no database.
enable_image_mirroring	true	Mirror the ntfy image into Artifact Registry.

Group 5 — Identity-Aware Proxy (IAP)

Variable	Default	Description
enable_iap	false	Require Google sign-in. Blocks unauthenticated publish/subscribe. Needs iap_oauth_client_id / _secret.

Group 6 — Environment Variables & Secrets

Variable	Default	Description
environment_variables	{}	Extra NTFY_* settings (e.g. NTFY_BASE_URL, NTFY_AUTH_DEFAULT_ACCESS).
secret_environment_variables	{}	Map of env var → Secret Manager secret name (optional; none required).

Group 6 — GKE Cluster

Variable	Default	Description
gke_cluster_name	""	Leave empty to auto-discover the Services_GCP cluster.
service_type	LoadBalancer	How the Service is exposed.
session_affinity	None	Set ClientIP to pin a client's stream to one pod if you scale replicas.

Group 7 — StatefulSet (durable message history)

Variable	Default	Description
stateful_pvc_enabled	null	Set true for a per-pod block PVC — durable message cache across restarts.
stateful_pvc_size	10Gi	Per-pod PVC storage size.
stateful_pvc_mount_path	/var/cache/ntfy	Mount path; matches NTFY_CACHE_FILE's directory.
stateful_pvc_storage_class	standard-rwo	StorageClass for the PVC.

Group 9 — Reliability Policies

Variable	Default	Description
enable_pod_disruption_budget	true	Protect availability during node upgrades.
pdb_min_available	1	Minimum pods available during voluntary disruptions.

Group 10 — Observability & Health

Variable	Default	Description
startup_probe_config	HTTP /v1/health	Startup probe. ntfy becomes healthy within seconds.
health_check_config	HTTP /v1/health	Liveness probe.
uptime_check_config	disabled	Optional Cloud Monitoring uptime check.

Group 13 / 14 — Filesystem & Cloud Storage

Variable	Default	Description
enable_nfs	false	Enable to back the SQLite cache with NFS for durable history (alternative to a PVC).
nfs_mount_path	/var/cache/ntfy	NFS mount path.
storage_buckets	[]	Not required — ntfy uses no object storage.

Group 16 — Database Backend

Variable	Default	Description
database_type	NONE	ntfy has no external database; leave NONE.
application_database_name / application_database_user	ntfy	Inert unless an external database is deliberately enabled.

Group 15 — Redis Cache

Variable	Default	Description
enable_redis	false	Not required — ntfy has no Redis dependency.

Group 19 — Custom Domain, Static IP & Networking

Variable	Default	Description
enable_custom_domain	true	Provision Ingress for custom hostnames + managed certificate.
application_domains	[]	Hostnames to serve.
reserve_static_ip	true	Stable external IP across redeploys.

All other inputs follow standard App_GKE behaviour.

---

5. Outputs

These values are returned on a successful deployment and are the quickest way to locate and explore the running resources.

Output	Description
service_name	Kubernetes Service name.
namespace	Namespace the workload runs in.
service_cluster_ip	In-cluster ClusterIP.
stage_service_cluster_ips	Map of ClusterIPs for stage-specific services.
service_external_ip	External LoadBalancer IP (when a static IP is reserved).
service_url	URL to reach ntfy.
database_instance_name / database_name / database_user	Database identifiers — empty for the default NONE engine.
database_password_secret / database_host / database_port	Database endpoint fields — unused for NONE.
storage_buckets	Created Cloud Storage buckets (none by default).
network_name / network_exists / regions	VPC network, presence, available regions.
container_image / container_registry	Deployed image and Artifact Registry repo.
monitoring_enabled / monitoring_notification_channels	Monitoring status and channels.
initialization_jobs / db_import_job	Names of any setup and import jobs (none by default).
deployment_id / tenant_id / resource_prefix	Naming identifiers.
project_id / project_number	Project identifiers.
cicd_enabled / cicd_configuration	CI/CD status and details (repo, trigger, registry).
github_repository_url / github_repository_owner / github_repository_name	CI/CD GitHub details.
artifact_registry_repository / cloudbuild_trigger_name / cloudbuild_trigger_id	Registry and build trigger.
kubernetes_ready	Whether the cluster/workload is ready.
vpc_sc_enabled / vpc_sc_perimeter_name / vpc_sc_dry_run_mode	VPC-SC status.
audit_logging_enabled / artifact_registry_cmek_enabled	Audit logging and CMEK status.

---

6. Configuration Pitfalls & Sensible Defaults

Risk: Critical (data loss / outage / security) — High (service degraded) — Medium (cost or partial degradation) — Low (minor).

Inherited plan-time validation. This module passes its configuration through the App_GKE foundation engine, which validates values and combinations at plan time — IAP with no OAuth credentials, enable_cloudsql_volume = true with database_type = "NONE", min_instance_count > max_instance_count, quota_memory_* without binary units. The GKE variant's own validation.tf enforces these guards. Invalid configuration fails the plan with a clear, named error before any resource is created, so most mistakes below are caught up front rather than at apply or runtime.

Setting	Sensible value	Risk	Consequence if wrong
stateful_pvc_enabled / enable_nfs (for durable history)	Enable one when history matters	High	With the default stateless Deployment and ephemeral cache, all message history is lost on every pod restart.
stateful_pvc_mount_path	/var/cache/ntfy	High	Mounting the PVC anywhere other than the NTFY_CACHE_FILE directory persists the wrong path and the cache stays ephemeral.
max_instance_count	1	High	Scaling beyond 1 splits subscribers across pods with no shared bus, so a message published to one pod is not delivered to subscribers on another.
enable_cloudsql_volume	false	High	Setting true with database_type = "NONE" starts an Auth Proxy sidecar with no instance to reach — rejected by the plan-time guard.
session_affinity	ClientIP if you scale	High	Without stickiness, a reconnecting subscriber lands on a different pod and misses cached messages held by the original pod.
enable_iap	only when auth-gated	High	IAP requires Google sign-in for every request, blocking unauthenticated publish/subscribe.
min_instance_count	1	High	GKE requires min ≥ 1; the validation guard rejects invalid values, and 0 would leave no pod to hold streams.
NTFY_BASE_URL	Actual external URL	Medium	Unset, attachment and web-push links resolve to the wrong host.
ntfy access control	Configure post-deploy	Medium	Left default, any client can publish to and subscribe from any topic on a public IP.
quota_memory_requests / _limits	binary units (4Gi, 8192Mi)	Critical	Bare integers are bytes and block all pod scheduling in the namespace.
enable_pod_disruption_budget	true	Medium	Disabling allows GKE to evict all pods simultaneously during maintenance.
application_version	Pin v2.x.y in prod	Low	latest maps to a pinned base (v2.11.0); pin explicitly to control upgrades.

---

For the foundation behaviour referenced throughout — IAM and Workload Identity, autoscaling, ingress and certificates, CI/CD, Cloud Armor, IAP, Binary Authorization, VPC-SC, backups, and image mirroring — see App_GKE. ntfy-specific application configuration shared with the Cloud Run variant is described in Ntfy_Common.

Related guides

• Hands-on lab: Ntfy on GKE Autopilot — deploy it step by step, with the console screens and commands at each stage.
• Ntfy on Google Cloud Run — the same application on Cloud Run, for when you need the other deployment target.
• Ntfy Common — Shared Application Configuration — the configuration shared by both deployment targets.