OnlyOffice on GKE Autopilot

ONLYOFFICE Document Server is an open-source collaborative online office suite for real-time co-editing of text documents, spreadsheets, presentations, PDFs, and forms — a self-hosted alternative to Google Docs / Microsoft Office Online. It is not usually opened directly by end users; instead it is embedded by a host application (Nextcloud, ownCloud, Seafile, or a custom integration) via its API and a shared JWT secret. This module deploys OnlyOffice 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 OnlyOffice 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 OnlyOffice costs on RAD, and how that compares

OnlyOffice on RAD's GKE Autopilot module costs about US$150–223 a month, with the module fee charged once in RAD credits. GKE Autopilot suits OnlyOffice that must stay up, scale across pods or run beside other Kubernetes workloads; for the lowest cost, the Cloud Run module runs the same OnlyOffice for about US$102 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 OnlyOffice, 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	300 credits (US$30 at the top-up price)	270 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$150 a month for the default configuration (table below) — more if this isn't your only Autopilot cluster	Metered hourly in credits; RAD publishes 112 credits a day, about 3,360 a month (about US$336 at the top-up price, US$269 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, 2x vCPU / 4 GiB	US$79.35
Cloud SQL for PostgreSQL, 1 vCPU / 3.75 GB, zonal, 10 GB SSD	US$51.02
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$150 (about US$223 if this isn't your only Autopilot cluster)

How it compares

Option	Price	Who runs it
ONLYOFFICE DocSpace, Startup	Free	ONLYOFFICE's own hosted cloud. Up to 3 admins, 2 GB storage, unlimited free users
ONLYOFFICE DocSpace, Business	$20/admin/month	ONLYOFFICE's own hosted cloud. 250 GB per admin, unlimited rooms
RAD, Cloud Run, your own project	About US$102, plus US$11 once per deployment	RAD's automation, in your project
RAD, Cloud Run, a RAD-managed project	About 960 credits (US$77–96)	RAD
RAD, GKE Autopilot	See the GKE guide	RAD

ONLYOFFICE's own DocSpace is priced per admin, not per editing user, which is the fairer comparison to RAD's flat infrastructure cost. For a small team that needs only one or two admins, DocSpace's free Startup tier or its $20/month Business plan can be cheaper in cash than RAD's own-project running cost — the difference is that this module deploys the Document Server you embed into your own application, rather than ONLYOFFICE's bundled workspace, which matters if you're already running a document platform (Nextcloud, ownCloud, a custom app) that only needs the editing engine.

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

If OnlyOffice 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 OnlyOffice 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 database, the VM, 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$150. Deploy OnlyOffice, 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 OnlyOffice 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 OnlyOffice 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 305 credits (US$30.50) in your own project, or 275 credits in a RAD-managed one. Deleting saves money once OnlyOffice would otherwise sit unused for about 6 days or more in your own project (about US$4.99 a day), or about 2 days or more in a RAD-managed one (112 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 OnlyOffice 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 OnlyOffice 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 OnlyOffice 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; ONLYOFFICE DocSpace pricing. Prices change; check each source before relying on a figure.

1. Overview

The onlyoffice/documentserver image is "batteries included": it bundles its own converters, nginx, and RabbitMQ (AMQP) under supervisord. This module externalizes PostgreSQL (Cloud SQL) and Redis; the bundled RabbitMQ stays internal on localhost.

Capability	Google Cloud service	Notes
Compute	GKE Autopilot	StatefulSet pods on port 80, 2 vCPU / 4Gi memory by default
Database	Cloud SQL for PostgreSQL 15	Required — only POSTGRES_13/14/15 (or NONE) pass a plan-time guard
Cache / editing state	External Redis	Mandatory — the bundled RabbitMQ stays internal, but session/editing state must be shared via an external Redis; defaults to the co-located NFS-VM Redis when redis_host is blank
Block storage	GKE Persistent Disk (standard-rwo)	20Gi PVC per pod at /var/www/onlyoffice/Data — the Document Server's cache/index data, which gcsfuse would corrupt
File persistence	Cloud Filestore (NFS)	Attachment/document storage shared across pods at /opt/onlyoffice/storage
Object storage	Cloud Storage	Declared (storage suffix) but not created by default (create_cloud_storage = false) — persistence lives on the block PVC
Secrets	Secret Manager	Auto-generated JWT_SECRET (48 characters); database password managed separately
Ingress	Cloud Load Balancing	External LoadBalancer with a reserved static IP; custom domain + managed certificate enabled by default

Sensible defaults worth knowing up front:

• PostgreSQL 15 is required. database_type defaults to POSTGRES_15; a plan-time guard rejects anything other than POSTGRES_13/14/15/NONE — MySQL is not supported.
• Redis is mandatory, not optional. A plan-time precondition fails the deployment if enable_redis = false. With redis_host left blank, enable_nfs must stay true so the NFS server IP can serve as the default Redis host.
• Data lives on a block PVC, not gcsfuse. stateful_pvc_enabled = true by default provisions a 20Gi standard-rwo (SSD) PVC per pod mounted at /var/www/onlyoffice/Data and auto-selects workload_type = "StatefulSet" — Document Server's caches/indexes under that path would corrupt on gcsfuse. create_cloud_storage is correspondingly false. NFS is enabled by default (enable_nfs = true, mounted at /opt/onlyoffice/storage) for shared attachment storage, and doubles as the fallback Redis host.
• Cloud SQL is reached via the Auth Proxy sidecar on loopback. A cloud-sql-proxy sidecar (enable_cloudsql_volume = true) listens on 127.0.0.1:5432; the wrapper cloud-entrypoint.sh maps the injected DB_* values onto Document Server's own variable names (DB_HOST/DB_PORT/DB_NAME/DB_USER already match; DB_PWD is set from DB_PASSWORD).
• JWT signing is on by default. JWT_ENABLED = "true" with a 48-character JWT_SECRET generated once and stored in Secret Manager — any host application embedding the editor must present the same secret.
• "latest" is pinned at build time. The custom build derives its base tag from an app-specific ONLYOFFICE_VERSION build ARG (not the Foundation-injected APP_VERSION, which the Dockerfile does not use); application_version = "latest" maps to 8.3.3.
• Scaling defaults to min=1, max=5, session_affinity = "ClientIP". Each StatefulSet pod has its own independent PVC — editing/session state is shared through Postgres and Redis, not the PVC.

---

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 OnlyOffice workload

OnlyOffice pods are scheduled on Autopilot, which bills for the CPU/memory the pods actually request. Because stateful_pvc_enabled = true by default, the workload is a StatefulSet with a 20Gi block PVC per pod.

• Console: Kubernetes Engine → Workloads → select the OnlyOffice workload for pods, revisions, and events. Kubernetes Engine → Services & Ingress shows the external IP.
• CLI:

  kubectl get pods,svc,statefulset -n "$NAMESPACE"
  kubectl logs -n "$NAMESPACE" statefulset/<service-name> --tail=100
  kubectl describe pod -n "$NAMESPACE" -l app=<service-name>

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

B. Cloud SQL for PostgreSQL 15

OnlyOffice stores document metadata, versions, and application state in a managed Cloud SQL for PostgreSQL 15 instance. Pods reach it through the Cloud SQL Auth Proxy sidecar on 127.0.0.1:5432; no public IP is exposed. On first deploy the db-init job creates the application role, database, and grants — the Document Server then installs its own schema on first boot.

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

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

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

C. External Redis, block storage (PVC) & Cloud Filestore (NFS)

Redis holds session/editing state that must be shared across every OnlyOffice replica — the bundled RabbitMQ stays internal, but Redis is externalized and mandatory (a plan-time guard fails the deployment if enable_redis = false). With redis_host left blank, the Foundation injects the co-located NFS-VM Redis IP (enable_nfs must stay true in that case); set redis_host explicitly to point at a different Redis instance. Two distinct persistence layers are wired in alongside it: a per-pod block PVC (standard-rwo, 20Gi by default) mounted at /var/www/onlyoffice/Data for caches/indexes/fonts (a real block device — gcsfuse would corrupt it), and Cloud Filestore (NFS) mounted at /opt/onlyoffice/storage, shared across all pods for attachment/document storage.

• Console: Compute Engine → VM instances (the NFS/Redis co-located VM); Kubernetes Engine → Storage (PVC/StorageClass); Filestore → Instances.
• CLI:

  kubectl exec -n "$NAMESPACE" statefulset/<service-name> -- env | grep REDIS_SERVER
  kubectl get pvc,sc -n "$NAMESPACE"
  gcloud filestore instances list --project "$PROJECT"

See App_GKE for how enable_redis/redis_host resolve to the NFS-VM IP, and the GKE SSD-vs-HDD stateful_pvc_storage_class trade-off.

D. Cloud Storage (optional)

A storage bucket is declared by OnlyOffice_Common but not created unless create_cloud_storage = true is set explicitly — by default all persistence lives on the block PVC and NFS above, not GCS.

• Console: Cloud Storage → Buckets.
• CLI:

  gcloud storage buckets list --project "$PROJECT" --filter="name~onlyoffice"

See App_GKE for CMEK options and GCS Fuse mounts.

E. Secret Manager

One OnlyOffice-specific secret is generated automatically and stored in Secret Manager: JWT_SECRET (48 characters, no special characters), which signs every internal Document Server API request and must be presented by any host application that embeds the editor. The database password is managed separately by the foundation. On GKE, secrets are projected into pods via the Secret Store CSI driver.

• Console: Security → Secret Manager.
• CLI:

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

See App_GKE for the Secret Store CSI integration and rotation.

F. Networking & ingress

By default the workload is exposed through an external Cloud Load Balancing IP (service_type = LoadBalancer, reserve_static_ip = true so the address survives redeploys), with enable_custom_domain = true allowing a Google-managed certificate once application_domains is populated.

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

  kubectl get svc,ingress -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. OnlyOffice Application Behaviour

• First-deploy database setup, no separate migration job. The db-init job runs using postgres:15-alpine. It resolves the Cloud SQL host (proxy sidecar on 127.0.0.1, falling back to the instance private IP), waits for PostgreSQL to be reachable, creates/updates the application role (LOGIN CREATEDB) with the generated password, creates the application database (owned by postgres, since the Cloud SQL superuser cannot SET ROLE to application roles), grants full privileges on the database and public schema, then signals the Auth Proxy to shut down so the Job pod completes. It only provisions the role/database/grants — the Document Server installs its own schema on first boot. The job is safe to re-run (execute_on_apply = true).
• JWT secret usage. JWT_ENABLED = "true", JWT_HEADER = "Authorization", JWT_IN_BODY = "true" are set by OnlyOffice_Common; the JWT_SECRET value itself is injected from Secret Manager (generated once, 48 characters). Every internal Document Server API call is signed with it, and any host application embedding the editor (Nextcloud, ownCloud, custom integration) must be configured with the identical secret — never rotate it after integrations are wired up.
• DB and Redis env-var mapping. cloud-entrypoint.sh runs before the upstream launcher: it sets DB_TYPE=postgres and DB_PWD from the injected DB_PASSWORD (DB_HOST/DB_PORT/DB_NAME/DB_USER already match Document Server's own names), and maps REDIS_HOST/REDIS_PORT/REDIS_AUTH onto REDIS_SERVER_HOST/REDIS_SERVER_PORT/REDIS_SERVER_PASS before exec-ing /app/ds/run-document-server.sh.
• Health path. Startup probe is HTTP GET /healthcheck (90s initial delay, 15s period, up to 40 failures — roughly 10 minutes of first-boot headroom while the bundled stack comes up and the schema installs). Liveness probe is the same path with a 120s delay / 30s period / 3 failures. /healthcheck returns true only once nginx and the document services are up and the database is reachable, and is served unauthenticated.
• Scaling constraints. min_instance_count = 1, max_instance_count = 5 by default; session_affinity = "ClientIP". Each StatefulSet pod has its own independent block PVC — replicas do not share the PVC, so scaling out is safe as long as Postgres and Redis (the shared state) are reachable by every pod.
• Inspect the init job and running config:

  kubectl get jobs -n "$NAMESPACE"
  kubectl logs -n "$NAMESPACE" job/<db-init-job-name>
  kubectl exec -n "$NAMESPACE" statefulset/<service-name> -- env | grep -E 'DB_|REDIS_SERVER|JWT_'

---

4. Configuration Variables

Variables are grouped exactly as they appear in variables.tf ({{UIMeta group=N}}). Only settings specific to or notable for OnlyOffice are listed; every other input is inherited from App_GKE with its standard behaviour and defaults.

Group 3 — Application Identity

Variable	Default	Description
application_name	onlyoffice	Base name for resources. Do not change after first deploy.
application_version	latest	onlyoffice/documentserver image tag; latest is pinned to 8.3.3 at build time via the ONLYOFFICE_VERSION build ARG.

Group 4 — Runtime & Scaling

Variable	Default	Description
container_port	80	The bundled nginx listens on port 80.
container_resources	{cpu_limit="2000m", memory_limit="4Gi"}	The bundled stack (Postgres client/Redis client/RabbitMQ/nginx/converters under supervisord) needs at least 4Gi.
min_instance_count / max_instance_count	1 / 5	Pod replica bounds.
enable_cloudsql_volume	true	Auth Proxy sidecar (loopback 127.0.0.1:5432) — required on GKE.

Group 6 — GKE Backend & Cluster

Variable	Default	Description
service_type	LoadBalancer	External IP for the OnlyOffice UI/API.
workload_type	null → StatefulSet	Auto-selected because stateful_pvc_enabled = true.
session_affinity	ClientIP	Sticky routing so a client reaches the same pod during a session.

Group 7 — StatefulSet Configuration

Variable	Default	Description
stateful_pvc_enabled	true	Document Server's cache/index data must live on a block PVC — gcsfuse would corrupt it. Auto-selects workload_type = "StatefulSet".
stateful_pvc_size	20Gi	Per-pod PVC size.
stateful_pvc_mount_path	/var/www/onlyoffice/Data	The Document Server data directory.
stateful_pvc_storage_class	standard-rwo	SSD-backed by default; override to standard (HDD) if the project's SSD_TOTAL_GB quota is tight.

Group 10 — Observability & Health

Variable	Default	Description
startup_probe	HTTP /healthcheck, 90s delay, 15s period, 40 failures	Generous first-boot budget — the bundled stack is slow to become ready.
liveness_probe	HTTP /healthcheck, 120s delay, 30s period, 3 failures	Restarts a wedged pod after boot.

Group 13 — Filesystem (NFS)

Variable	Default	Description
enable_nfs	true	Shared attachment storage; also the default Redis host source when redis_host is blank.
nfs_mount_path	/opt/onlyoffice/storage	Where OnlyOffice stores shared attachments.

Group 14 — Cloud Storage

Variable	Default	Description
create_cloud_storage	false	Off by default — persistence lives on the block PVC and NFS, not GCS. Set true only if you need an additional plain bucket.

Group 15 — Redis

Variable	Default	Description
enable_redis	true	Required — a plan-time guard rejects false.
redis_host	"" (→ NFS server IP)	Leave blank to use the co-located NFS-VM Redis; enable_nfs must be true in that case.
redis_port	6379	Redis TCP port.

Group 16 — Database Backend

Variable	Default	Description
database_type	POSTGRES_15	Required — a plan-time guard rejects anything but POSTGRES_13/14/15/NONE. MySQL is not supported.
application_database_name	onlyoffice	Database name. Immutable after first deploy.
application_database_user	onlyoffice	Application database user; password auto-generated in Secret Manager.

Group 19 — Custom Domain, Static IP & Networking

Variable	Default	Description
enable_custom_domain	true	Provisions a Kubernetes Ingress for custom domain routing (enabled by default, unlike most apps).
reserve_static_ip	true	Stable external IP across redeploys.
application_domains	[]	Custom hostnames + managed certificate.
network_tags	["nfsserver"]	Required tag for the default NFS/Redis co-location path.

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 OnlyOffice.
database_instance_name	Cloud SQL instance name.
database_name / database_user	Application database name / user.
database_password_secret	Secret Manager secret holding the DB password.
database_host / database_port	DB endpoint (127.0.0.1 via the Auth Proxy) / port.
storage_buckets	Created Cloud Storage buckets (empty unless create_cloud_storage = true).
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 (db-init) 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 (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 — a StatefulSet forced alongside a stateless setting, IAP with no authorized identities, quota_memory_* given as bare integers, an out-of-range container_port/backup_retention_days. OnlyOffice additionally layers its own guards (PostgreSQL-only database_type, mandatory enable_redis, redis_host/enable_nfs coupling). 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
database_type	POSTGRES_15 (or 13/14)	Critical	Any other engine is rejected at plan time — MySQL is not supported by Document Server.
enable_redis	true	Critical	A plan-time guard rejects false — without shared Redis, session/editing state cannot be coordinated across pods.
redis_host / enable_nfs	Leave redis_host blank only with enable_nfs = true	Critical	Blank redis_host with enable_nfs = false fails at plan time — there is no Redis host to resolve.
JWT_SECRET (auto-generated)	Never change after integrations exist	Critical	Rotating it breaks every host application (Nextcloud/ownCloud/etc.) embedding the editor until all are updated with the new value.
application_database_name / application_database_user	Set once	Critical	Immutable after first deploy; renaming recreates the DB/user and orphans all data.
stateful_pvc_enabled	true	High	Disabling it (or forcing workload_type = "Deployment" alongside it) risks gcsfuse-corrupted cache/index data — or fails at plan time if forced.
stateful_pvc_storage_class	standard-rwo (SSD) or standard (HDD) under tight quota	Medium	SSD draws the tight SSD_TOTAL_GB quota; a wide campaign of stateful apps can exhaust it — see App_GKE.
enable_cloudsql_volume	true	High	The Auth Proxy sidecar on 127.0.0.1:5432 is required for DB connectivity on GKE.
container_resources.memory_limit	4Gi	High	The bundled Postgres/Redis-client/RabbitMQ/nginx/converter stack under supervisord is heavy; undersizing risks OOM during startup.
max_instance_count	5 (tune to load)	Medium	Each pod's converter workload is CPU/memory intensive; scaling too high without headroom risks node pressure under Autopilot.
quota_memory_requests / _limits	binary units (4Gi, 8192Mi)	Critical	Bare integers are treated as bytes and block all pod scheduling in the namespace.
reserve_static_ip	true	Medium	Without it, the external IP can change across redeploys, breaking DNS and any registered integration callback URL.
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. OnlyOffice-specific application configuration shared with the Cloud Run variant is described in OnlyOffice_Common.

Related guides

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