Castopod on GKE Autopilot

Castopod is an open-source, ActivityPub-native podcast hosting platform built on CodeIgniter 4 (PHP 8) and served by FrankenPHP/Caddy. This module deploys Castopod 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 Castopod 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

Castopod runs as a single FrankenPHP/Caddy web workload. The deployment wires together a focused set of Google Cloud services:

Capability	Google Cloud service	Notes
Compute	GKE Autopilot	FrankenPHP/Caddy pods on port 8080, 1 vCPU / 2 GiB by default
Database	Cloud SQL for MySQL 8.0	Required — the engine is fixed at MYSQL_8_0; Castopod does not support PostgreSQL
File persistence	Cloud Filestore (NFS)	Podcast media (audio, artwork) persists under /var/lib/castopod, shared across pods
Object storage	Cloud Storage	Two buckets are provisioned by default (suffixes data and media) — neither is mounted into the pod unless gcs_volumes is configured
Cache	Redis (optional)	Castopod defaults to a filesystem cache (CP_CACHE_HANDLER = file); Redis is opt-in
Secrets	Secret Manager	Auto-generated CP_ANALYTICS_SALT; database password
Ingress	Cloud Load Balancing / Gateway API	External LoadBalancer with a reserved static IP and zero-config HTTPS via an auto-issued <ip>.nip.io hostname

Sensible defaults worth knowing up front:

• MySQL 8.0 is mandatory. The database engine is fixed by the shared application layer (the variant passes database_type = null, which keeps the Common default MYSQL_8_0); other engines are not supported and break the CodeIgniter migrations.
• Cloud SQL is reached via the Auth Proxy sidecar on loopback, but the app dials it over TCP. enable_cloudsql_volume = true mounts a cloud-sql-proxy sidecar listening on 127.0.0.1:3306, and the Common module hardcodes DB_HOST = 127.0.0.1 for the service container. Because Castopod (CodeIgniter 4) reads its database connection from dot-notated database.default.* keys — which cannot be expressed as Cloud Run/K8s env var names — the platform entrypoint writes them into Castopod's .env file at container start rather than injecting them as env vars.
• Two replicas of configuration exist for health probes — only one is effective. Castopod's app-specific startup_probe/liveness_probe variables (Group 10) always win; the generic health_check_config/startup_probe_config variables inherited from App_GKE are forwarded but structurally overridden by the Common module's own probe config, so they have no effect for this application.
• Single replica by default. min_instance_count = 1, max_instance_count = 1. The NFS-backed workload deploys with the Recreate strategy, so do not scale beyond 1 without verifying shared-storage behaviour for media uploads and the object cache.
• NFS is enabled by default (enable_nfs = true, mounted at /var/lib/castopod) so uploaded episode audio and artwork persist across pod restarts and are shared across replicas — Castopod stores media on the filesystem, not in the database.
• Session affinity is ClientIP so a client's requests reach the same pod.
• Zero-config HTTPS is on by default. enable_custom_domain = true and reserve_static_ip = true are both defaults; with application_domains left empty, App_GKE provisions a Gateway and derives a free <ip>.nip.io hostname with a Google-managed certificate — no DNS setup required to reach Castopod over HTTPS.
• CP_ANALYTICS_SALT is generated automatically and stored in Secret Manager. It anonymises podcast listener analytics and must stay stable after first boot.
• No separate migration job. The castopod/castopod image runs the CodeIgniter 4 schema migrations automatically on every container start, so the schema is created on first boot once the db-init job has provisioned the database and user.
• First-run setup is manual. After deploy, open the service URL and complete Castopod's web install wizard to create the first super-admin account and configure the instance/podcast defaults.

---

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 Castopod workload

Castopod pods are scheduled on Autopilot, which bills for the CPU/memory the pods actually request. Because the workload is NFS-backed, the Deployment uses the Recreate strategy (a rolling update would run two pods against the same NFS-backed media directory and shared DB and deadlock).

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

  kubectl get pods,svc -n "$NAMESPACE" --selector="app~castopod" 2>/dev/null || kubectl get pods,svc -n "$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 MySQL 8.0

Castopod stores all application data (podcasts, episodes, users, analytics) in a managed Cloud SQL for MySQL 8.0 instance. Pods reach it through the Cloud SQL Auth Proxy sidecar on 127.0.0.1:3306; no public IP is exposed. On first deploy the db-init job creates the application database, user, and grants; the CodeIgniter migrations then create the schema on the app container's first boot.

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

  gcloud sql instances list --project "$PROJECT" --filter="name~castopod"
  gcloud sql instances describe <instance-name> --project "$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. Cloud Storage & file persistence

Two Cloud Storage buckets are provisioned automatically by default — a generic data bucket (the App_GKE foundation default) and a Castopod-specific media bucket declared by Castopod_Common. Neither is mounted into the pod filesystem unless gcs_volumes is explicitly configured; instead, Castopod's actual media directory (/var/www/castopod/public/media) is persisted via NFS (Cloud Filestore) mounted at /var/lib/castopod, shared across pods.

• Console: Cloud Storage → Buckets; Filestore → Instances.
• CLI:

  gcloud storage buckets list --project "$PROJECT" --filter="name~castopod"
  gcloud filestore instances list --project "$PROJECT"
  kubectl get pvc -n "$NAMESPACE"

See App_GKE for CMEK options and GCS Fuse mounts.

D. Redis (optional object cache)

Redis is disabled by default — Castopod uses a filesystem cache (CP_CACHE_HANDLER = file). When enable_redis = true, the module injects REDIS_HOST/REDIS_PORT for Castopod's object cache; if redis_host is left empty, the foundation resolves it to the NFS server VM's IP (requires enable_nfs = true).

• Console: Memorystore → Redis (if using a managed instance instead of the NFS-colocated Redis).
• CLI:

  kubectl exec -n "$NAMESPACE" deploy/<service-name> -- env | grep -i redis

E. Secret Manager

One cryptographic secret is generated automatically and stored in Secret Manager: CP_ANALYTICS_SALT (used to anonymise podcast listener analytics). 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 "$PROJECT" --filter="name~analytics-salt"
  gcloud secrets versions access latest --secret=<secret-name> --project "$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). Because enable_custom_domain = true by default, App_GKE additionally provisions a Gateway and — when application_domains is left empty — a zero-config <ip>.nip.io hostname with a Google-managed certificate, so Castopod is reachable over HTTPS immediately without owning a domain. A real custom domain can be supplied instead.

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

  kubectl get svc,gateway,httproute -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. Castopod Application Behaviour

• First-deploy database setup. The db-init job runs db-init.sh using mysql:8.0-debian. It connects to Cloud SQL (Unix socket under /cloudsql via the Auth Proxy sidecar, falling back to private-IP TCP), idempotently creates the application database, user, and grants, verifies the app user can connect, then shuts down the proxy sidecar. The job is safe to re-run (execute_on_apply = true, max_retries = 3).
• Migrations run on container start (no separate migration job). The castopod/castopod image runs the CodeIgniter 4 schema migrations automatically on every startup, so the schema is created on first boot once db-init has provisioned the database and user, and upgrading application_version applies schema changes without a separate job.
• Database config lives in .env, materialised at container start. Castopod (CodeIgniter 4) reads its default connection from framework-native, dot-notated keys (database.default.hostname|database|username|password|port|DBDriver|DBPrefix) that cannot be expressed as Kubernetes env var names. The platform wrapper entrypoint writes them into Castopod's .env from the foundation-injected DB_HOST (which on GKE is 127.0.0.1, the Auth Proxy sidecar) and DB_NAME/DB_USER/DB_PASSWORD, then delegates to the upstream FrankenPHP/Caddy entrypoint.
• CP_BASEURL is derived automatically. When not explicitly set, the entrypoint derives it from the foundation-injected GKE_SERVICE_URL and writes it as app.baseURL in .env, so podcast feed and media links reflect the real service address (including the auto-issued nip.io hostname).
• CP_ANALYTICS_SALT should be stable after first boot. It is generated once and written to Secret Manager; changing it breaks de-duplication continuity for previously recorded analytics, though it does not corrupt existing rows.
• NFS-backed rollouts use Recreate. Updates terminate the old pod before starting the new one, avoiding two pods deadlocking on the shared NFS-backed media directory and DB locks.
• Health path. The startup probe is TCP on the container port with a 30-second initial delay and a 20-retry window (period_seconds = 15), giving first-boot CodeIgniter migrations ample time to complete. The liveness probe is HTTP GET / with a 300-second (5-minute) initial delay — Castopod's unauthenticated homepage returns 200 once booted and connected to MySQL.
• First-run setup. After deploy, open the service URL and complete Castopod's web install wizard to create the first super-admin account and set the instance name and podcast defaults. Media uploads then persist to the NFS-backed media directory.
• 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" deploy/<service-name> -- cat /var/www/castopod/.env

---

4. Configuration Variables

Variables are grouped exactly as they appear on the deployment platform (by their {{UIMeta group=N}} tag, which does not always match the section headings in the source file). Only settings specific to or notable for Castopod 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	castopod	Base name for resources. Do not change after first deploy.
application_version	latest	castopod/castopod image tag used as the custom-build base; latest is pinned to a known-good tag (1.15.5) at build time via the app-specific CASTOPOD_VERSION build ARG.

Group 4 — Runtime & Scaling

Variable	Default	Description
cpu_limit	1000m	1 vCPU per Castopod container instance.
memory_limit	2Gi	Minimum ~512Mi to boot; 2Gi recommended for large media libraries.
min_instance_count	1	Keep at 1 to keep the workload reachable.
max_instance_count	1	Keep at 1 unless shared NFS/cache behaviour is verified for multiple pods.
container_port	8080	Castopod's FrankenPHP/Caddy server listens on 8080.
enable_cloudsql_volume	true	Auth Proxy sidecar (loopback) — required on GKE.
php_memory_limit	512M	PHP memory limit; raise for memory-intensive plugins.
upload_max_filesize / post_max_size	64M	Max upload / POST size; keep post_max_size ≥ upload_max_filesize for episode audio uploads.

Group 6 — GKE Backend & Cluster

Variable	Default	Description
service_type	LoadBalancer	External IP for the Castopod UI and public podcast feeds.
workload_type	null → Deployment	Deployment (NFS-backed, Recreate strategy).
session_affinity	ClientIP	Sticky routing so a client reaches the same pod.

Group 10 — Observability & Health

Variable	Default	Description
startup_probe	TCP, 30s delay, 20 retries	Effective startup probe — supersedes the generic startup_probe_config, which is forwarded but overridden and has no effect.
liveness_probe	HTTP /, 300s delay	Effective liveness probe against Castopod's unauthenticated homepage — supersedes the generic health_check_config.

Group 13 — Filesystem (NFS)

Variable	Default	Description
enable_nfs	true	NFS is on by default so uploaded episode audio/artwork persist and are shared.
nfs_mount_path	/var/lib/castopod	Where Castopod's shared media state is mounted.

Group 15 — Redis Cache

Variable	Default	Description
enable_redis	false	Switches Castopod's object cache from the filesystem to Redis.
redis_host	""	Redis endpoint. Leave empty to use the NFS server IP (requires enable_nfs = true).
redis_port	6379	Redis port.

Group 16 — Database Backend

Variable	Default	Description
database_type	null → MYSQL_8_0	Keeps the Common MySQL 8.0 default; Castopod does not support other engines.
application_database_name	castopod	Database name. Immutable after first deploy.
application_database_user	castopod	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 Gateway; with application_domains empty, yields a zero-config <ip>.nip.io HTTPS hostname.
application_domains	[]	Set to use a real custom hostname + managed certificate instead of the nip.io fallback.
reserve_static_ip	true	Stable external IP across redeploys — required for the nip.io hostname to stay constant.

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 Castopod.
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 (data and media).
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. 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	null (→ MYSQL_8_0)	Critical	Selecting a non-MySQL engine breaks the CodeIgniter migrations and every DB-backed route.
application_database_name / application_database_user	Set once	Critical	Immutable after first deploy; renaming recreates the DB/user and destroys all podcast data.
enable_nfs	true	Critical	With NFS off, uploaded episode audio and artwork live on ephemeral disk and are lost on every pod restart/redeploy.
CP_ANALYTICS_SALT (auto-generated)	Never change	High	Changing it after first boot breaks listener de-duplication continuity for previously recorded analytics.
enable_cloudsql_volume	true	High	The Auth Proxy sidecar on 127.0.0.1:3306 is required for DB connectivity on GKE.
max_instance_count	1	High	Scaling beyond 1 without verified shared-storage/cache behaviour risks inconsistent media state and duplicated analytics.
session_affinity	ClientIP	High	Without stickiness, requests bounce between pods and disrupt authenticated admin sessions.
memory_limit	2Gi	High	Below ~512Mi the PHP/FrankenPHP pod fails to boot; large media libraries need more headroom.
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, both the external IP and any auto-derived nip.io hostname can change across redeploys, breaking bookmarked/RSS feed URLs.
enable_custom_domain + empty application_domains	Fine as default (nip.io)	Low	Produces a working but non-branded <ip>.nip.io URL; set application_domains for a real hostname.
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. Castopod-specific application configuration shared with the Cloud Run variant is described in Castopod_Common.