Dify on GKE Autopilot

Dify is an open-source LLM application development platform for building production-grade AI applications with a visual workflow builder, RAG pipeline, agent framework, multi-model management, and built-in observability. This module deploys Dify 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 Dify uses and how to explore and operate them from the Google Cloud Console and the command line. For the mechanics that are 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 Dify costs on RAD, and how that compares

Dify on RAD's GKE Autopilot module costs about US$149.81 a month in a Google Cloud project you own, plus a one-time 300-credit module fee (US$30). Dify is a visual builder for production LLM applications, with a RAG pipeline, agent framework and built-in observability. For the lowest cost, the Cloud Run guide runs the same Dify for about US$102 a month; pick GKE here only if Dify must stay up continuously, scale across pods, or run beside other Kubernetes workloads.

If Dify is only needed occasionally — a demo, a trial, a seasonal or study workload — a RAD-managed project can be deleted and restored within 30 days for a handful of credits, so that US$149.81 becomes a few dollars a month instead. See Pause it for free, below.

What you pay on RAD

	In a Google Cloud project you own	In a project RAD manages for you
Module fee, once per deployment	300 credits (US$30 at the top-up price)	270 credits (10% lower)
Build time	A few credits per build (roughly 3–6)	The same
Google Cloud running cost	Billed by Google to your own billing account, about US$149.81 a month for the default configuration (table below)	Metered hourly in credits; RAD publishes 116 credits a day, about 3480 a month (about US$348 at the top-up price)
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, less on a 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 database.

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

Resource (module default)	Per month
GKE Autopilot pod, 2x vCPU / 4 GiB	US$79.35
Cloud SQL for PostgreSQL, 1 vCPU / 3.75 GB, zonal, 10 GB SSD	US$51
NFS/cache file server (small VM)	US$13.43
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$149.81

How it compares

• The project's own managed offering, Dify Cloud, has a free Sandbox plan, then $590 per workspace a year (about $49 a month) on Professional and $1,590 per workspace a year (about $133 a month) on Team.
• Run it yourself and a VPS running Dify's API, worker and web containers plus PostgreSQL yourself costs less in cash: sized to roughly 2 vCPU/4 GiB, a comparable VPS (Hetzner or DigitalOcean) runs about $24–48 a month, or a GCP e2-standard-2-class VM about $49 — plus your own time for backups, security patches and upgrades, which RAD's managed Cloud SQL, Secret Manager and monitoring cover for you.
• RAD never claims to be cheaper than bare infrastructure in cash terms — it usually is not. The difference is what RAD manages for you: patching, backups, secrets and monitoring, plus the ability to pause Dify for free rather than paying for idle capacity (see below).
• Most of the running cost above is the managed PostgreSQL database (and, where used, the shared file server) — the part a bare VPS comparison has to run and back up by hand.

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

If Dify 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 Dify that you only need occasionally 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 Cloud SQL, the VM or GKE Autopilot resources one by one — the whole project, and everything in it, simply stops.
• Restoring costs a handful of 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 is a handful of credits (under US$1) in total for a typical Dify chain, against the 305 credits a full redeploy costs.
• So a month of occasional use can cost a few dollars, not US$149.81. Deploy Dify, use it, 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 hours it 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: the backup bucket does not survive. Nightly 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 have customised Dify 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

GKE Autopilot keeps at least one pod running, so it never scales to zero on its own. 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 305 credits (US$30.50) in your own project. Deleting saves money only once Dify would otherwise sit unused for about 7 days or more in your own project (US$5 a day), or about 3 days or more in a RAD-managed one (116 credits a day).
• Most of the running cost is usually the database and the shared file or cache server. They stop only when nothing else in the project uses them, so deleting Dify while something else shares the project saves only this app's own compute part.
• Keep your data first. Nightly backups are written 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.
• Note your settings. RAD does not recreate a deleted deployment for you; you enter the settings again when you deploy.

Lab sessions and Managed Environments

• Lab sessions, for training. A trainer runs a session for a class. Each participant gets the app 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 the app 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; RAD's own fees and daily-credit estimates from radmodules.dev/pricing; Delete and restore projects for the 30-day recovery window and what it says about Cloud Storage objects without soft delete; Dify Cloud. Prices change; check each source before relying on a figure.

1. Overview

Dify runs as a Python/Flask API container (with an embedded Celery worker under supervisord) plus a separate Next.js web frontend. The deployment wires together a focused set of Google Cloud services:

Capability	Google Cloud service	Notes
Compute	GKE Autopilot	API+worker pod (2 vCPU / 4 GiB by default) + web frontend pod, horizontally autoscaled
Database	Cloud SQL for PostgreSQL 15	Required — pgvector extension enabled for vector storage
Vector store	pgvector (in-database)	Reuses the Cloud SQL instance; no separate vector database needed
Shared files	Filestore (NFS)	Shared Redis host co-location; NFS VM also used for task state
Object storage	Cloud Storage	A dedicated gcs-dify<resource-prefix>-storage bucket for uploaded files and assets
Cache & task queue	Redis	Required for Celery broker/backend and SSE/WebSocket LLM streaming
Secrets	Secret Manager	Auto-generated SECRET_KEY and database password
Ingress	Cloud Load Balancing	External LoadBalancer, optional custom domain + managed certificate

Sensible defaults worth knowing up front:

• PostgreSQL 15 is mandatory. MySQL and NONE are not supported; Dify requires PostgreSQL for all metadata, workflow state, and user accounts.
• pgvector is always enabled. The vector extension is installed on the Cloud SQL instance automatically, making the same database instance the vector store — no extra service required.
• Redis is required. Celery (workflow execution, document indexing, async LLM calls) and the SSE/WebSocket event bus both depend on Redis. Disabling it breaks all background processing.
• NFS is enabled by default. The NFS server VM hosts the Redis process when no external Redis host is set.
• A web frontend is deployed automatically. A langgenius/dify-web Deployment is wired to the API service URL — you do not need to configure it separately.
• SECRET_KEY is auto-generated and stored in Secret Manager; it signs Dify sessions and must never be changed after first deployment.
• Database migrations run on every pod start (via MIGRATION_ENABLED=true), so version upgrades apply schema changes automatically.

---

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 Dify workload

Dify runs two Autopilot Deployments: the API+worker pod (Flask/gunicorn + Celery via supervisord) and the web frontend (Next.js). Horizontal Pod Autoscaling sizes each Deployment between the minimum and maximum replica counts.

• Console: Kubernetes Engine → Workloads → select the Dify API or web workload to see pods, events, and resource usage. Kubernetes Engine → Services & Ingress shows the external IP.
• CLI:

  kubectl get pods,svc,hpa -n "$NAMESPACE"
  kubectl logs -n "$NAMESPACE" deploy/<service-name> --tail=100
  kubectl describe hpa -n "$NAMESPACE"

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

B. Cloud SQL for PostgreSQL 15

Dify stores all application data (workflows, knowledge bases, user accounts, API keys) in a managed Cloud SQL for PostgreSQL 15 instance. Pods connect privately through the Cloud SQL Auth Proxy sidecar over a Unix socket — no public IP is exposed. On first deploy an initialization Job creates the application database and user. The pgvector extension is installed automatically so the same instance serves as the vector store.

• Console: SQL → select the instance for connections, backups, flags, and metrics.
• CLI:

  gcloud sql instances list --project "$PROJECT"
  gcloud sql instances describe <instance-name> --project "$PROJECT"
  gcloud sql connect <instance-name> --user=<db-user> --project "$PROJECT"

The instance name, database name, user, and the Secret Manager secret holding the password are all surfaced in the Outputs. For the connection model, automated backups, and password rotation, see App_GKE.

C. Filestore (NFS) and Cloud Storage

A Filestore (NFS) share is mounted into every pod. The NFS server VM also runs the Redis process used as the Celery broker when no external Redis host is configured. A dedicated Cloud Storage bucket (gcs-dify<resource-prefix>-storage) is provisioned for uploaded files and assets; Dify's google-storage driver accesses it via Workload Identity — no service account key file is needed.

• Console: Filestore → Instances for the NFS share; Cloud Storage → Buckets.
• CLI:

  gcloud filestore instances list --project "$PROJECT"
  gcloud storage buckets list --project "$PROJECT"
  gcloud storage ls gs://<dify-storage-bucket>/
  kubectl exec -n "$NAMESPACE" deploy/<service-name> -- df -h | grep -i nfs

See App_GKE for NFS provisioning, GCS Fuse, and CMEK options.

D. Redis — Celery and event bus

Redis is required for three functions in Dify:

Role	Redis DB	Purpose
Celery broker & backend	db 1	Queues and tracks all background tasks (LLM inference, document indexing)
Event bus	db 0	SSE/WebSocket streaming for real-time LLM output
General cache	db 0	Application caching

When no external Redis host is configured, the NFS server VM IP is used as the Redis endpoint. For production, point redis_host at a dedicated Memorystore for Redis instance.

• Console: Memorystore → Redis (if using a managed instance).
• CLI:

  redis-cli -h <redis-host> ping
  redis-cli -h <redis-host> info keyspace

E. Secret Manager

The Dify SECRET_KEY (used for JWT signing and session encryption) and the database password are stored as Secret Manager secrets and injected into pods at runtime; plaintext never appears in pod specs. The SECRET_KEY is generated once and must not be rotated while the deployment is running — all pods must share the same value.

• Console: Security → Secret Manager.
• CLI:

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

The database password secret name is in the Outputs. See App_GKE for the Secret Store CSI integration and rotation.

F. Networking & ingress

By default the Dify service is exposed through an external Cloud Load Balancing IP. A custom domain with a Google-managed certificate can be enabled, and a static IP can be reserved so the address survives redeploys.

• 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.

G. Cloud Logging & Monitoring

Pod stdout/stderr flow to Cloud Logging; GKE and Cloud SQL metrics flow to Cloud Monitoring. Optional uptime checks and alert policies are available.

• 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. Dify Application Behaviour

• First-deploy database setup. An initialization Job (db-init) connects to Cloud SQL via the Auth Proxy and idempotently creates the Dify database user and database. It runs automatically and is safe to re-run.
• Migrations on start. Each pod runs Dify's Flask-Migrate database migrations on startup (MIGRATION_ENABLED=true), so upgrading the application version applies schema changes automatically. No separate migration job is needed.
• API + worker in one pod. The custom container wraps langgenius/dify-api with supervisord. Both the gunicorn API server and the Celery worker run inside the same pod — they share CPU and memory allocation. Size accordingly: 2 vCPU and 4 GiB is the recommended minimum.
• Web frontend. A langgenius/dify-web Deployment is deployed automatically and wired to the API service URL. Access Dify through the web service external IP or the web_url output.
• LLM provider API keys. Provider keys (OpenAI, Anthropic, etc.) are configured per-workspace via the Dify web console and stored in the application database. Use secret_environment_variables only for environment-level configuration that cannot be set in the UI.
• CORS. WEB_API_CORS_ALLOW_ORIGINS and CONSOLE_CORS_ALLOW_ORIGINS default to "*". Restrict to your domain via environment_variables in production.
• Health path. Readiness and liveness probes target /health with a 30-second initial delay to allow the API server and database migrations to complete on first boot.

---

4. Configuration Variables

Variables are grouped exactly as they appear on the deployment platform. Only settings specific to or notable for Dify 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 — Deployment Environment

Variable	Default	Description
tenant_id	demo	Short suffix that makes resource names unique per environment.
support_users	[]	Emails granted project access and monitoring alerts.
resource_labels	{}	Labels applied to all resources for cost and ownership tracking.

Group 3 — Application Identity

Variable	Default	Description
application_name	dify	Base name for resources. Do not change after first deploy.
display_name	Dify - LLM Application Platform	Friendly name shown in the Console.
description	(set)	Workload description annotation.
application_version	0.15.0	Dify image version tag; applies to both API and web containers. Pin to a specific version in production.

Group 4 — Runtime & Scaling

Variable	Default	Description
deploy_application	true	Set false to provision infrastructure only.
cpu_limit	2000m	CPU per pod; 2 vCPU minimum — gunicorn and Celery share this allocation.
memory_limit	4Gi	Memory per pod; 4 GiB recommended for LLM workflow caching and document processing.
min_instance_count	1	Minimum replicas. Keep ≥ 1 so the Celery worker maintains its Redis broker connection.
max_instance_count	3	Maximum replicas (autoscaler ceiling).
container_port	5001	Dify API server listens on port 5001.
enable_cloudsql_volume	true	Cloud SQL Auth Proxy sidecar for Unix socket connections. Required for database connectivity.
enable_vertical_pod_autoscaling	false	VPA tunes resource requests automatically; enabling it disables HPA.
enable_pod_disruption_budget	false	Creates a PodDisruptionBudget for minimum availability during node maintenance.
pdb_min_available	1	Minimum pods available during disruptions.

Group 5 — Environment Variables & Secrets

Variable	Default	Description
environment_variables	{}	Extra non-secret settings. Use to override WEB_API_CORS_ALLOW_ORIGINS, LOG_LEVEL, etc.
secret_environment_variables	{}	Map of env var → Secret Manager secret name. Use for LLM provider API keys.

Group 6 — GKE Backend & Cluster

Variable	Default	Description
service_type	LoadBalancer	How the Service is exposed.
session_affinity	ClientIP	Sticky routing — recommended for Dify's session state.
workload_type	null	Auto-resolves to StatefulSet when per-pod storage is enabled.
network_tags	['nfsserver']	Node/pod tags; nfsserver is required for NFS connectivity.
gke_cluster_name	""	GKE cluster name. Leave empty to auto-discover.
namespace_name	""	Kubernetes namespace. Auto-generated when empty.
termination_grace_period_seconds	60	Seconds Kubernetes waits after SIGTERM before force-killing.

Group 7 — StatefulSet Configuration

Variable	Default	Description
stateful_pvc_enabled	null	Enable per-pod PVCs. Setting true auto-selects StatefulSet workload type.
stateful_pvc_size	10Gi	Storage size for each PVC.
stateful_pvc_mount_path	/data	Container path where the per-pod PVC is mounted.
stateful_pvc_storage_class	standard-rwo	Kubernetes StorageClass for PVC provisioning.

Group 10 — Observability & Health

Variable	Default	Description
startup_probe_config / startup_probe	HTTP /health, 30 s delay	Startup probe — container receives no traffic until /health returns 200.
health_check_config / liveness_probe	HTTP /health	Liveness probe.
uptime_check_config	disabled	Optional Cloud Monitoring uptime check.
alert_policies	[]	Optional metric alert policies.

Group 11 — Jobs & Scheduled Tasks

Variable	Default	Description
initialization_jobs	[]	Leave empty to use the built-in db-init job. Provide a non-empty list to replace it entirely.
cron_jobs	[]	Scheduled Kubernetes CronJobs.
additional_services	[]	Additional Kubernetes Deployments alongside Dify (web frontend is wired automatically).

Group 12 — CI/CD & GitHub Integration

Standard App_GKE Cloud Build / Cloud Deploy integration — see App_GKE. Key inputs: enable_cicd_trigger, github_repository_url, github_token, enable_cloud_deploy, enable_binary_authorization.

Group 13 — Filesystem (NFS)

Variable	Default	Description
enable_nfs	true	Shared Filestore volume; also provides the default Redis host when no external Redis is set.
nfs_mount_path	/mnt/nfs	Mount path inside the container.
nfs_instance_name	""	Name of an existing NFS GCE VM. Leave empty to auto-discover.
nfs_instance_base_name	app-nfs	Base name for an inline NFS GCE VM when none exists.

Group 14 — Cloud Storage & Artifact Registry

Variable	Default	Description
create_cloud_storage	true	Provision additional GCS buckets. The gcs-dify<resource-prefix>-storage bucket is always provisioned by Dify_Common.
storage_buckets	[{ name_suffix = "data" }]	Additional buckets beyond the auto-provisioned storage bucket.
gcs_volumes	[]	GCS Fuse mounts via the CSI driver.
manage_storage_kms_iam / enable_artifact_registry_cmek	false	CMEK options.

Group 16 — Database Backend

Variable	Default	Description
db_name	dify_db	PostgreSQL database name. Immutable after first deploy.
db_user	dify_user	Application user. Immutable after first deploy.
database_password_length	32	Generated password length (16–64).
enable_auto_password_rotation	false	Zero-downtime DB password rotation.

Group 17 — Backup & Maintenance

Variable	Default	Description
backup_schedule	0 2 * * *	Automated backup cron (UTC).
backup_retention_days	7	Retention; raise to 30–90 for production/compliance.
enable_backup_import / backup_source / backup_uri	restore options	Restore from a backup on deploy.

Group 18 — Custom SQL Scripts

enable_custom_sql_scripts, custom_sql_scripts_bucket, custom_sql_scripts_path, custom_sql_scripts_use_root — run SQL from a GCS bucket after provisioning. See App_GKE.

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. Do not set false for Dify — see the Critical pitfall in §6: the web frontend's API calls resolve to an unreachable internal hostname without it.

Group 20 — Identity-Aware Proxy (IAP)

Variable	Default	Description
enable_iap	false	Require Google sign-in in front of Dify. Requires enable_custom_domain = true.
iap_authorized_users / iap_authorized_groups	[]	Who may access.
iap_oauth_client_id / iap_oauth_client_secret	""	Required when IAP is enabled (sensitive).

Group 21 — Redis Cache

Variable	Default	Description
enable_redis	true	Required. Enables Redis for Celery task queue and SSE/WebSocket streaming.
redis_host	""	Leave empty to use the NFS server IP; set explicitly for an external Memorystore instance.
redis_port	6379	Redis port.
redis_auth	""	Optional Redis auth password (sensitive).

Group 22 — VPC Service Controls & Audit Logging

Variable	Default	Description
enable_vpc_sc	false	Enforce a VPC-SC perimeter (requires organization_id).
vpc_cidr_ranges / vpc_sc_dry_run	(set)	Access level CIDRs / dry-run mode.
enable_audit_logging	false	Detailed Cloud Audit Logs.

---

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 (Dify API).
namespace	Namespace the workload runs in.
service_cluster_ip	In-cluster ClusterIP of the API service.
stage_service_cluster_ips	Map of ClusterIPs for stage-specific services (Cloud Deploy).
service_external_ip	External LoadBalancer IP (when a static IP is reserved).
web_url	URL of the Dify web frontend (use this to open the browser UI).
database_instance_name	Cloud SQL instance name.
database_name	Application database name.
database_user	Application database user.
database_password_secret	Secret Manager secret holding the DB password.
database_host / database_port	DB endpoint (127.0.0.1 via Auth Proxy) / port.
storage_buckets	Created Cloud Storage buckets.
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 the setup and (optional) import jobs.
deployment_id / tenant_id / resource_prefix	Naming identifiers.
project_id / project_number	Project identifiers.
cicd_enabled / cicd_configuration	CI/CD status and details.
github_repository_url / github_repository_owner / github_repository_name	CI/CD repository details.
artifact_registry_repository / cloudbuild_trigger_name / cloudbuild_trigger_id	Registry and build trigger.
kubernetes_ready	Whether the cluster and workload are 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).

Setting	Sensible value	Risk	Consequence if wrong
enable_redis	true (required)	Critical	All Celery tasks (workflow execution, document indexing, async LLM calls) fail silently without Redis.
enable_cloudsql_volume	true (required)	Critical	The Auth Proxy sidecar is the only path to PostgreSQL; disabling it breaks all database connectivity.
SECRET_KEY (auto-generated)	immutable once set	Critical	All pods must share the same key; rotating it logs out all users and invalidates active sessions.
db_name / db_user	set once	Critical	Immutable after first deploy; renaming recreates the database and destroys data.
enable_backup_import	false unless restoring	Critical	Enabling without a valid backup_uri fails the import job.
enable_redis + enable_nfs	both true if no external Redis	Critical	Without NFS, there is no Redis host when redis_host is empty — Celery fails to start.
secret_environment_variables for LLM keys	always use secret refs	Critical	Plain env vars expose API keys in pod specs visible via kubectl describe pod.
enable_redis + redis_host	correct host	High	Incorrect redis_host produces a malformed Celery broker URL; all async tasks queue indefinitely.
reserve_static_ip + service_type	true / LoadBalancer — do not change for Dify	Critical	The web frontend's CONSOLE_API_URL/APP_API_URL resolve via the $(GKE_SERVICE_URL) sentinel to local.service_url, which falls back to the unreachable internal *.svc.cluster.local hostname if reserve_static_ip=false or service_type is overridden away from LoadBalancer. The page loads, but every browser-side API call fails with net::ERR_NAME_NOT_RESOLVED. This is a genuine exception to the fleet-wide reserve_static_ip=false IP-quota-conservation convention.
memory_limit	4Gi	High	Too little memory causes OOM kills during document ingestion or LLM workflow caching.
min_instance_count	1	High	Scale-to-zero causes cold starts and abandons in-flight Celery tasks.
timeout_seconds	300 (raise for workflows)	High	Multi-step workflows and RAG indexing can exceed 300 s; increase to 3600 for complex deployments.
WEB_API_CORS_ALLOW_ORIGINS	restrict in production	High	Default "*" allows cross-origin requests from any domain.
application_version	pin to a specific version	Medium	Unpinned versions risk unexpected schema migrations that break the application on redeploy.
enable_iap / enable_cloud_armor	enable for production	Medium	The Dify console is publicly reachable without these controls.
backup_retention_days	7 (raise for prod)	Medium	Too short for compliance retention.

---

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. Dify-specific application configuration shared with the Cloud Run variant is described in Dify_Common.

Related guides

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