infra/stacks/nextcloud/chart_values.yaml
Viktor Barzin fb1e47a20a nextcloud: re-enable Keel auto-upgrades with occ-upgrade self-heal + live-tag floor
Re-enrolls Nextcloud in Keel (opted out after the 2026-05-26 32.0.3->32.0.9
bump stuck the pod in maintenance mode ~22h). Two safeguards engineer around
both failure modes:

- F1 (interrupted occ upgrade -> 503): nextcloud-watchdog CronJob runs
  `occ upgrade` + clears maintenance mode when occ reports needsDbUpgrade=true;
  Job deadline bumped 120->600s so it isn't killed mid-migration.
- F2 (helm re-renders a tag below the Keel-bumped live image -> downgrade
  CrashLoop): chart_values renders the live tag via a plural
  kubernetes_resources data source (empty-list-on-absence -> floor 32.0.9 on
  fresh install/DR), so a re-render never downgrades below live.

Scope is patch -- Kyverno's shared inject-keel-annotations policy stamps it and
its background-controller overrides a TF-set value, and patch == minor for
Nextcloud in practice (32.0.x only; major 33 stays manual). Dropped the
per-workload keel.sh/policy override resources to avoid perpetual drift; ns
enrollment + Kyverno now own the keel annotations like other workloads.

Also bumps the external-storage bootstrap Job create timeout 1m->12m to match
its own 10m pod-wait, since Keel bumps now roll the pod mid-apply.

Verified: Keel auto-upgraded 32.0.9->32.0.10 on apply, entrypoint occ upgrade
completed clean (no watchdog needed), pod 2/2, HTTP 200, plan shows no drift.
2026-06-01 19:50:41 +00:00

191 lines
5.5 KiB
YAML

# image.tag is rendered dynamically (templatefile var `image_tag`) from the
# CURRENT live Deployment tag, falling back to var.nextcloud_image_tag_floor
# (32.0.9) on fresh install / DR — see stacks/nextcloud/main.tf
# `data.kubernetes_resource.nextcloud_live` + locals. This makes helm upgrades
# image-no-ops in steady state and means a re-render can NEVER downgrade below
# the Keel-bumped live tag (the 2026-06-01 CrashLoop: a pinned 32.0.3 lost to
# live 32.0.9 and Nextcloud refused the downgrade). Keel (keel.sh/policy=minor)
# bumps the live tag upward within major 32; the next apply just follows it.
# flavor=apache renders the bare apache-default tag (live image is
# `nextcloud:<tag>`, no -apache suffix).
image:
flavor: apache
tag: "${image_tag}"
nextcloud:
host: nextcloud.viktorbarzin.me
trustedDomains:
- nextcloud.viktorbarzin.me
# mail:
# enabled: true
# # the user we send email as
# fromAddress: nextcloud@viktorbarzin.me
# # the domain we send email from
# domain: viktorbarzin.me
# smtp:
# host: mail.viktorbarzin.me
# secure: starttls
# port: 587
# authtype: LOGIN
# name: nextcloud@viktorbarzin.me
# password:
extraEnv:
- name: TRUSTED_PROXIES
value: "10.0.0.0/8"
- name: PHP_MEMORY_LIMIT
value: "512M"
- name: PHP_UPLOAD_LIMIT
value: "16G"
# - name: mail_smtpdebug
# value: "true"
# - name: loglevel
# value: "0"
configs:
zzz-redis.config.php: |
<?php
// Redis via HAProxy master-only service. HAProxy (3 replicas, PDB
// minAvailable=2) health-checks all v2 pods via `INFO replication` and
// routes to the current role:master. Sentinel failover takes <30s, and
// HAProxy detects the new master via its 1s tcp-check interval and
// starts routing within ~3s of detection. Removed the old in-process
// sentinel-query loop on 2026-04-19 after the Redis rework — see
// beads code-v2b and infra/docs/architecture/databases.md.
$CONFIG = array(
'memcache.distributed' => '\\OC\\Memcache\\Redis',
'memcache.locking' => '\\OC\\Memcache\\Redis',
'redis' => array(
'host' => 'redis-master.redis.svc.cluster.local',
'port' => 6379,
'password' => '',
'timeout' => 1.5,
'read_timeout' => 1.5,
),
);
performance.config.php: |
<?php
$CONFIG = array(
'loglevel' => 2,
// Cap + rotate nextcloud.log. Without this it grew unbounded to
// 10GB+ and bloated every backup (2026-06-01 space incident).
// At 10MB the log rotates to nextcloud.log.1 (1 kept) → ~20MB max.
'log_rotate_size' => 10485760,
'mail_smtpdebug' => false,
);
zzz-mysql.config.php: |
<?php
$CONFIG = array(
'mysql.utf8mb4' => true,
);
phpConfigs:
zzz-custom.ini: |
max_execution_time = 300
max_input_time = 300
default_socket_timeout = 300
opcache.enable_file_override = 1
apc.shm_size = 128M
extraVolumes:
- name: apache-tuning
configMap:
name: nextcloud-apache-tuning
- name: db-password-sync
configMap:
name: nextcloud-db-password-sync
defaultMode: 0755
- name: pve-nfs
persistentVolumeClaim:
claimName: nextcloud-pve-nfs-root
- name: pve-nfs-ssd
persistentVolumeClaim:
claimName: nextcloud-pve-nfs-ssd-root
extraVolumeMounts:
- name: apache-tuning
mountPath: /etc/apache2/mods-available/mpm_prefork.conf
subPath: mpm_prefork.conf
- name: db-password-sync
mountPath: /docker-entrypoint-hooks.d/before-starting
- name: pve-nfs
mountPath: /mnt/pve-nfs
- name: pve-nfs-ssd
mountPath: /mnt/pve-nfs-ssd
internalDatabase:
enabled: false
externalRedis:
enabled: false
externalDatabase:
enabled: true
type: mysql
host: ${mysql_host}
user: nextcloud
database: nextcloud
existingSecret:
secretName: nextcloud-db-creds
usernameKey: db-username
passwordKey: DB_PASSWORD
persistence:
enabled: true
existingClaim: nextcloud-data-encrypted
accessMode: ReadWriteOnce
size: 20Gi
startupProbe:
enabled: true
initialDelaySeconds: 30
periodSeconds: 10
timeoutSeconds: 30
failureThreshold: 60
successThreshold: 1
livenessProbe:
enabled: true
initialDelaySeconds: 30
periodSeconds: 60
timeoutSeconds: 30
failureThreshold: 10
successThreshold: 1
readinessProbe:
enabled: true
initialDelaySeconds: 30
periodSeconds: 60
timeoutSeconds: 30
failureThreshold: 5
successThreshold: 1
podAnnotations:
diun.enable: "true"
diun.include_tags: "^[0-9]+(?:.[0-9]+)?(?:.[0-9]+)?.*"
dependency.kyverno.io/wait-for: "mysql.dbaas:3306,redis-master.redis:6379"
secret.reloader.stakater.com/reload: "nextcloud-db-creds"
# OnRootMismatch: kubelet only recursively chowns the volume to fsGroup if the
# root dir's GID doesn't already match. Without this, every pod restart triggers
# a ~30-min recursive chown of /srv/nfs and /srv/nfs-ssd (600k+ files) — the
# default policy "Always" recurses every time. Locks fsGroup=33 explicitly so
# this block fully replaces the chart's default {fsGroup: 33}.
securityContext:
fsGroup: 33
fsGroupChangePolicy: OnRootMismatch
collabora:
enabled: false # Using onlyoffice instead
resources:
limits:
memory: 8Gi
requests:
cpu: 50m
memory: 256Mi
cronjob:
enabled: true
resources:
limits:
memory: 384Mi
requests:
cpu: 25m
memory: 384Mi