EvolutionAPI on GKE Autopilot

Evolution API is an open-source Node.js WhatsApp Business API gateway (built on the Baileys library) that provisions WhatsApp instances, sends and receives messages, and exposes a REST API plus a manager UI for wiring WhatsApp into other systems. This module deploys Evolution API 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 Evolution API 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.

1. Overview

Evolution API runs as a Node.js web workload on GKE Autopilot. The deployment wires together a focused set of Google Cloud services:

Capability	Google Cloud service	Notes
Compute	GKE Autopilot	Node.js pod, 2 vCPU / 4 GiB by default; pinned to a single replica
Database	Cloud SQL for PostgreSQL 15	Required — Evolution API uses Prisma against PostgreSQL only
Cache	Redis	Enabled by default (`CACHE_REDIS_URI`); caches instance/message state
Object storage	Cloud Storage	A dedicated data bucket provisioned automatically
Secrets	Secret Manager	Auto-generated `AUTHENTICATION_API_KEY`; database password
Ingress	Cloud Load Balancing	External LoadBalancer with `ClientIP` session affinity; optional custom domain + managed certificate

Sensible defaults worth knowing up front:

PostgreSQL 15 is mandatory. The database engine is fixed by the shared application layer; Evolution API uses Prisma and does not support other engines.
The workload is pinned to a single replica (min_instance_count = 1, max_instance_count = 1). Evolution API holds live WhatsApp (Baileys) socket sessions in memory, per-pod; those sessions are not shared across replicas, so scaling out fragments live connections. Do not raise max_instance_count.
AUTHENTICATION_API_KEY is generated automatically and stored in Secret Manager. It is Evolution API's global admin key and must never be rotated after first boot — rotating it makes already-provisioned WhatsApp instances unreachable and returns 401 to every client still holding the old key.
Redis is enabled by default (enable_redis = true). Leave redis_host empty to use the NFS server VM's IP as the Redis endpoint (requires enable_nfs = true), or point it at an explicit managed instance.
Session affinity is ClientIP and the Service type is LoadBalancer by default, exposing an external IP for WhatsApp webhook callbacks and the manager UI.
Memory defaults to 4 GiB — Evolution API needs at least 2 GiB for reliable operation.
SERVER_URL is defaulted at runtime by the container entrypoint from the injected service URL, so QR-code and webhook callback URLs reflect the real address.

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 Evolution API workload

Evolution API pods are scheduled on Autopilot, which bills for the CPU/memory the pods actually request. The workload runs as a single replica by design.

Console: Kubernetes Engine → Workloads → select the Evolution API workload to see pods, revisions, 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 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

Evolution API stores all application data (WhatsApp instances, contacts, chats, message history) in a managed Cloud SQL for PostgreSQL 15 instance. Pods reach it privately through the Cloud SQL Auth Proxy sidecar over a TCP loopback (127.0.0.1); no public IP is exposed. On first deploy an initialization Job creates the application database and user; Prisma migrations then create the schema on container boot.

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"
# Open an interactive shell to inspect schema/data:
gcloud sql connect <instance-name> --user=evolution --database=evolution --project "$PROJECT"

The instance name, database (evolution), user (evolution), 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. Cloud Storage

A dedicated Cloud Storage data bucket is provisioned automatically for Evolution API file storage. The workload service account is granted access. Additional buckets can be declared via storage_buckets.

Console: Cloud Storage → Buckets.

CLI:

gcloud storage buckets list --project "$PROJECT"
gcloud storage ls gs://<data-bucket>/          # bucket name is in the Outputs

See App_GKE for CMEK options and GCS Fuse mounts.

D. Redis (cache)

Redis is enabled by default (enable_redis = true). Evolution API uses it for instance/message caching (CACHE_REDIS_URI, Redis DB index 6). When redis_host is left empty and enable_nfs = true, the NFS server VM's IP is used as the Redis endpoint; the container entrypoint assembles the URI at runtime.

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

CLI:

redis-cli -h <redis-host> ping
redis-cli -h <redis-host> info keyspace
# Confirm the cache URI is injected into the running pod:
kubectl exec -n "$NAMESPACE" deploy/<service-name> -- env | grep CACHE_REDIS

E. Secret Manager

One cryptographic secret is generated automatically and stored in Secret Manager: AUTHENTICATION_API_KEY — Evolution API's global admin API key, mounted via the Secret Store CSI driver. The database password is managed separately by the foundation.

Console: Security → Secret Manager.

CLI:

gcloud secrets list --project "$PROJECT" --filter="name~api-key"
gcloud secrets versions access latest --secret=<secret-name> --project "$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 workload is exposed through an external Cloud Load Balancing IP with ClientIP session affinity — required so a WhatsApp client's requests stick to the single pod holding its socket. 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. The entrypoint emits [cloud-entrypoint] markers that confirm the resolved DB/Redis/URL config on boot. 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. EvolutionAPI Application Behaviour

First-deploy database setup. An initialization Job runs db-init.sh using postgres:15-alpine. It connects through the Cloud SQL Auth Proxy and idempotently creates the evolution database and role, grants privileges, and makes the app user the owner of the public schema, then signals the proxy sidecar to shut down so the Job pod completes. The job is safe to re-run.
Prisma migrations on start. Evolution API runs prisma migrate deploy on every container boot (via the wrapped deploy_database.sh), so upgrading the application version applies schema changes without a separate migration step. On GKE the cloud-sql-proxy sidecar is a TCP loopback, so the entrypoint connects with sslmode=disable.
AUTHENTICATION_API_KEY is immutable after first boot. The global admin key is generated once and written to Secret Manager. Rotating it makes all already-provisioned WhatsApp instances unreachable and returns 401 to every client still holding the old key. Only rotate during a planned migration.
Single-replica by design. WhatsApp (Baileys) socket sessions live in the pod's memory and are not shared across replicas. min_instance_count = 1 keeps one running; max_instance_count = 1 prevents fragmenting live connections. Do not scale out. ClientIP session affinity keeps each client pinned to the pod.
Webhook / QR callback URL. The entrypoint defaults SERVER_URL to the injected service URL. After the LoadBalancer external IP is assigned, set SERVER_URL (via environment_variables) to the external URL so QR-code and webhook callbacks use a reachable address:
```
kubectl get svc <service-name> -n "$NAMESPACE" \
  -o jsonpath='{.status.loadBalancer.ingress[0].ip}'
```
First-run setup. After deploy, retrieve AUTHENTICATION_API_KEY from Secret Manager and use it (as the apikey header) to reach the manager UI at /manager, create a WhatsApp instance (POST /instance/create), then scan the returned QR code from WhatsApp on your phone to connect the number.
Health path. Startup and liveness probes target the root / — an unauthenticated status endpoint that responds once the server is up. Allow several minutes on first boot while Prisma migrations run.

4. Configuration Variables

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