## Why
None of the 8 CNPG Postgres clusters in this repo had **any** backup configured. A
lost PVC, a fat-fingered migration, or a bad app deploy meant permanent, unrecoverable
data loss for authentik, litellm, artifactapi, woodpecker, puppet, encapi, paperclip
and grafana. This adds continuous WAL archiving + a nightly base backup to Ceph RGW for
every cluster, plus code-forward restore docs.
## What
- **`spec.backup.barmanObjectStore`** on each `cnpg_cluster.yaml` — turns on continuous
WAL archiving to `s3://cnpg-<app>`, WAL compressed with zstd, base backups with bzip2,
30-day retention. TLS to `s3.ceph.unkin.net` is trusted via the reflected
`vault-ca-cert` (`endpointCA`).
- **`cnpg_backup.yaml`** per app — a cephrgw `ObjectStoreUser` + `Bucket` (operator
provisions the bucket and mints the S3 key into `cnpg-<app>-backup-s3`; **nothing is
hardcoded**) and a staggered nightly `ScheduledBackup`.
- **`schemas/ceph.unkin.net/*.json`** — the three cephrgw CRD schemas so kubeconform can
validate the new CRs.
- **`docs/`** — new docs folder (README index + `cnpg-backups.md` + `cnpg-restore.md`).
## Design decisions (answers to the open questions)
**One bucket for all, or per-database?** → **Per-database (one bucket + owner user per
cluster).** The cephrgw CRDs are namespace-scoped (`BucketRef`/`OwnerRef` resolve only
*within the same namespace*), and CNPG reads its S3 credential Secret from its *own*
namespace. A single shared bucket would require either cross-namespace bucket refs
(unsupported) or hand-copying the S3 secret into all 8 namespaces (defeats "operator
mints the keys"). Per-namespace `s3://cnpg-<app>` with a dedicated owner user is the
simplest correct topology and needs zero manual seeding. Each user owns exactly one
bucket, so owner-level (full) access is already tightly scoped — no extra `BucketAccess`
grant needed.
**Backup mechanism.** The deployed CNPG operator is **v1.28** (helm chart
`cloudnative-pg-0.27.0`, appVersion 1.28.0). 1.26+ deprecates the in-tree
`barmanObjectStore` in favour of the Barman Cloud Plugin, but the plugin is **not
deployed**, and `barmanObjectStore` is still fully functional on 1.28. So this uses the
in-tree mechanism. Migrating to the plugin is a follow-up (noted in `docs/cnpg-backups.md`).
## Schedule / retention (defaults — Ben to adjust)
| App | Cluster | Bucket | Nightly base backup |
| --- | --- | --- | --- |
| authentik | postgres | cnpg-authentik | 01:00 |
| litellm | litellm-postgres | cnpg-litellm | 01:20 |
| artifactapi | postgres | cnpg-artifactapi | 01:40 |
| woodpecker | woodpecker-postgres | cnpg-woodpecker | 02:00 |
| puppet | puppet-postgres | cnpg-puppet | 02:20 |
| encapi | postgres | cnpg-encapi | 02:40 |
| paperclip | paperclip-postgres | cnpg-paperclip | 03:00 |
| grafana | postgres | cnpg-grafana | 03:20 |
Retention is **30d** across the board — flagged as a default to tune per cluster.
Schedules are staggered 20 min apart so 8 base backups don't hit RGW at once.
## Validation
- `kustomize build --enable-helm` + `kubeconform` (repo CI args, incl. the new ceph
schemas) pass on all 8 affected overlays (paperclip validated at base — it has no
overlay yet). ceph CRs resolve their schemas (`Skipped: 0`).
- `pre-commit run` passes on all changed files (yamllint, no-plain-secrets, etc.).
- Note: a full `ci/validate-apps.sh` run aborts locally on the unrelated
`cattle-system` overlay (`chart requires kubeVersion < 1.35 vs host helm v1.36.0`) —
pre-existing, reproduces on `origin/main`, unrelated to this change.
## Notes / caveats
- No overlap with the woodpecker chart-bump PR (#297, overlay files only) or the logging
PR (#296) beyond the three **identical** generated `schemas/ceph.unkin.net/*.json`
files, which merge cleanly whichever lands first.
- Credentials: no manual seeding — the cephrgw-operator mints the RGW user + keys. The
only prerequisite is the operator being healthy (it is, in `cephrgw-system`).
## Follow-ups
- Barman Cloud Plugin migration (deploy plugin, move clusters to `ObjectStore` CRs).
- Tune per-cluster retention / schedule if the defaults don't fit.
https://claude.ai/code/session_015ur3i7D2azsMAWTSVABApv
Reviewed-on: #298
Co-authored-by: Ben Vincent <ben@unkin.net>
Co-committed-by: Ben Vincent <ben@unkin.net>
## Why
Turns on right-sizing telemetry for the whole estate. Adds a `VerticalPodAutoscaler` with `updateMode: "Off"` (recommendation-only, advise mode) for every Deployment and StatefulSet in `apps/base`. Off mode never evicts or mutates pods, so this is purely observational: the VPA recommender (added in the vpa-system PR) publishes suggested requests/limits in each VPA's status, and nothing acts on them until someone deliberately flips a mode.
## Changes
- Add one `vpa.yaml` per app under `apps/base/<app>/` containing a `<workload>-vpa` VerticalPodAutoscaler for each workload, and register it in that app's `kustomization.yaml`.
- Coverage: 17 workloads across 11 apps.
- age-api (age-api), artifactapi (api, redis, ui), authentik (redis), bind-system (bind-operator), cephrgw-system (cephrgw-operator), encapi (encapi), kanidm (kanidm StatefulSet), litellm (litellm, redis), paperclip (paperclip), pdbmux (pdbmux), puppet (puppetboard, puppetdb, puppetserver-compiler, puppetserver-master).
## Skipped (intentionally)
- **CNPG `Cluster` objects** (artifactapi, authentik, encapi, grafana, litellm, paperclip, puppet, woodpecker) — Postgres is managed by CloudNativePG, not a VPA target.
- **CronJobs** (puppet g10k/generate-types, reposync x4) — not VPA-able.
## HPA / VPA caveat
api, ui (artifactapi), litellm, and all four puppet deployments also carry an HPA. With `updateMode: "Off"` there is no conflict today (VPA only recommends). VPA objects targeting these carry an inline comment: do **not** flip to `Auto`/`Initial` while the HPA still autoscales on CPU/memory, or the two controllers will fight over the same resource. Move the HPA to a custom/non-resource metric first.
## Verification
- `kubectl kustomize` over every touched overlay: 9/11 overlays PASS rendering all their VPAs; 2 failures are pre-existing and unrelated to this change — `authentik` fails on a local helm-tooling flag error identically on origin/main, and `paperclip` has no `apps/overlays/au-syd1/paperclip` directory yet. Both apps' base kustomizations build clean and render their VPA.
## Merge gate
- **Requires the vpa-system CRDs PR (argocd-apps #281) to merge first.** These manifests use `autoscaling.k8s.io/v1 VerticalPodAutoscaler`; without the CRDs installed, ArgoCD sync fails on an unknown kind.
Reviewed-on: #283
Co-authored-by: Ben Vincent <ben@unkin.net>
Co-committed-by: Ben Vincent <ben@unkin.net>
## Why
The puppet-on-k8s compilers classify nodes with a uv/python ENC script (`encapi-enc`). Each fresh compiler pod resolves the script's python dependencies on first invocation, and that resolution fails on cold pods (observed exits 135/2), breaking puppet agent catalog compilation. `encapic` (git.unkin.net/unkin/encapic) is a stdlib-only Go replacement with no runtime dependency resolution — a behavioural drop-in whose output matches the python script byte-for-byte.
## Changes
- Points the compiler `external_nodes` at `/opt/bin/encapic`.
- Reworks the `setup-shared-bins` init container to `curl` the encapic `v0.1.0` `encapic_linux_amd64` release binary (sha256-verified against the published `.sha256`, installed mode 0755) into the shared bins dir, instead of copying the python script and installing uv.
- Removes the `puppet-encapi-enc` configmap generator, its volume and mount, and the `resources/encapi-enc` script. uv was consumed solely by that script (grep of `apps/base/puppet` confirms no other consumer), so its installation is removed too.
`kubectl kustomize apps/overlays/au-syd1/puppet` builds clean.
## Merge gate
Do not merge until the encapic `v0.1.0` release assets exist:
`https://git.unkin.net/unkin/encapic/releases/download/v0.1.0/encapic_linux_amd64` (+ `.sha256`). The init container pulls them at pod start.
## Rollback
Revert this PR to restore the `encapi-enc` configmap script + uv install and repoint `external_nodes`.
---------
Co-authored-by: benvin <neotheo@gmail.com>
Reviewed-on: #277
Co-authored-by: Ben Vincent <ben@unkin.net>
Co-committed-by: Ben Vincent <ben@unkin.net>
## Why
The k8s puppetserver compilers classify nodes via an exec ENC that today queries legacy Cobbler (`https://cobbler.main.unkin.net`) over TLS. `encapi` now runs in-cluster and exposes a cobbler-wire-compatible endpoint (`GET /cblr/svc/op/puppet/hostname/<certname>`), a drop-in for the Cobbler URL. This cuts the puppet-on-k8s ENC over from Cobbler to encapi — a prerequisite for migrating VM agents onto puppet-on-k8s.
## Changes
- Rename the ENC script `resources/cobbler-enc` -> `resources/encapi-enc`, and its configmap `puppet-cobbler-enc` -> `puppet-encapi-enc` (kustomization configMapGenerator + deployment volume, initContainer copy path, and volumeMount subPath).
- Point `external_nodes` in the compiler `puppet.conf` at `/opt/bin/encapi-enc`.
- Target the in-cluster encapi service `http://encapi.encapi.svc.cluster.local` (plain HTTP), overridable via the `ENCAPI_URL` env var.
- Drop the `/opt/vault-ca-cert.crt` verify for the ENC request (no TLS needed in-cluster).
- Leave the response normalization identical: classes coerced to a list, `enc_role`/`enc_env` params set, `environment` stripped when it equals `testing`.
Verified with `kubectl kustomize apps/overlays/au-syd1/puppet` (builds clean, exit 0); the generated `puppet-encapi-enc` configmap contains the new URL and env var.
## 🚨 Merge gate
**Do not merge until encapi is seeded** (terraform-incus `benvin/encapi-seed` PR applied). An empty encapi means every node resolves to a 404. On 404 the ENC script exits non-zero, so puppet fails the compile rather than classifying the node with zero classes — nodes will fail to run until they exist in encapi. Seed encapi first so real nodes classify correctly; only unknown nodes should 404.
Reviewed-on: #272
Co-authored-by: Ben Vincent <ben@unkin.net>
Co-committed-by: Ben Vincent <ben@unkin.net>
## Summary
- Removes `clusterIP: null` from the `puppetdb` Service spec
## Why
Setting `clusterIP: null` makes ArgoCD's desired state explicit about the field being null. Kubernetes assigns a real IP on creation and the field is immutable afterward. The null vs assigned-IP mismatch causes permanent OutOfSync on the puppetdb Service. Removing the field means ArgoCD no longer claims ownership of `clusterIP`, so the API server's value is authoritative.
Reviewed-on: #166
## Summary
- Adds `group: gateway.networking.k8s.io` and `kind: Gateway` to all `parentRefs` entries
- Adds `group: ""`, `kind: Service`, and `weight: 1` to all `backendRefs` entries
- Affects 9 HTTPRoute files across artifactapi, cattle-system, consul, kanidm, litellm, paperclip, puppet, and vault
## Why
ArgoCD diffs the desired manifest against the live Kubernetes object. The Gateway API controller defaults these fields when creating/updating objects, so the live state always has them — causing persistent OutOfSync for every HTTPRoute. Same root cause as #153 (certificateRefs).
## Test plan
- [ ] All affected ArgoCD applications show Synced after merge
Reviewed-on: #162
The Gateway API admission server defaults certificateRefs[].group to ""
when it is omitted. ArgoCD diffed the desired state (no group field) against
the live state (group: "") and flagged every gateway as out of sync.
Fix: explicitly set group: "" in all certificateRefs entries so the
rendered manifest matches the API server's canonical form exactly.
Affected: artifactapi, cattle-system, consul, litellm, paperclip,
puppet (puppetboard + puppetdb), vault.
Reviewed-on: #153
Add port 80 HTTP listener and redirect HTTPRoute to artifactapi,
cattle-system (rancher), litellm, paperclip, and puppetboard — restoring
the redirect behaviour that existed on the previous nginx/traefik Ingress
resources.
Reviewed-on: #145
## Summary
- Replace `Ingress` (nginx) with `Gateway` + `HTTPRoute` using `traefik-internal` GatewayClass
- TLS terminated at the Gateway listener; cert-manager provisions the certificate via `vault-issuer`
- external-dns annotations moved to the Gateway
- `ingress_puppetboard.yaml` is unchanged in this PR (separate PR)
## Test plan
- [ ] ArgoCD syncs the puppet app cleanly
- [ ] cert-manager issues the `puppetdb-tls` certificate
- [ ] external-dns creates the DNS record
- [ ] `https://puppetdb.k8s.syd1.au.unkin.net` is reachable
Reviewed-on: #131
## Summary
- Replace `Ingress` (nginx) with `Gateway` + `HTTPRoute` using `traefik-internal` GatewayClass
- TLS terminated at the Gateway listener; cert-manager provisions the certificate via `vault-issuer`
- external-dns annotations moved to the Gateway
- `ingress_puppetdb.yaml` is unchanged in this PR (separate PR)
## Test plan
- [ ] ArgoCD syncs the puppet app cleanly
- [ ] cert-manager issues the `puppetboard-tls` certificate
- [ ] external-dns creates the DNS record
- [ ] `https://puppetboard.k8s.syd1.au.unkin.net` is reachable
Reviewed-on: #130
The g10k-code cronjob was failing with "Permission denied" because the
container (running as uid 999, non-root) attempted to create /shared in
the container root filesystem, which is not writable. Clone to /tmp
which is always writable by unprivileged users.
Reviewed-on: #76
The RWO puppetserver-shared-config PVC caused multi-attach errors when
the cronjob pod was scheduled on a different node than the previous run,
stalling the init container indefinitely. Since the config only needs to
exist for the duration of the job, remove the init container and PVC
entirely and clone the r10k config directly into /shared within the main
container before running g10k.
Reviewed-on: #75
The container was OOMKilled on every run because the 256Mi limit was far
too low for `puppet generate types`. Remove PUPPETSERVER_JAVA_ARGS (only
relevant to the puppetserver JVM, not the puppet CLI) and raise the
memory limit to 1Gi / request 512Mi.
Reviewed-on: #74
filemapper is not available on RubyGems under that name and was causing
puppetserver-compiler to crash loop. The interfaces provider that
requires puppetx/filemapper is Debian-specific and should not be loaded
on RedHat-based puppetservers.
Reviewed-on: #72
The network module's interfaces provider requires puppetx/filemapper
which was not installed, causing catalog compilation failures with
"no such file to load -- puppetx/filemapper".
Adds filemapper to additional-ruby-gems.sh for puppetserver/compiler
pods, installs it directly in the generate-types cronjob (which has no
access to that script), and adds cronjob_generate-types.yaml to the
kustomization so the CronJob is actually deployed.
💘 Generated with Crush
Assisted-by: Claude Sonnet 4.6 via Crush <crush@charm.land>
Reviewed-on: #71
Puppetboard was connecting to PuppetDB on port 8080 (plain HTTP), causing
403 Forbidden errors on the /metrics/v2 Jolokia endpoint which requires
HTTPS with a Puppet certificate. Also replaced the invalid
PUPPETDB_SSL_SKIP_VERIFY var with the correct PUPPETDB_SSL_VERIFY,
PUPPETDB_CERT, and PUPPETDB_KEY pointing to the certs already generated
by the cert-generator init container.
💘 Generated with Crush
Assisted-by: Claude Sonnet 4.6 via Crush <crush@charm.land>
Reviewed-on: #70
PuppetDB requires a separate read-only database user for its read pool.
Without it, it refuses to use the write user for read queries and all
/pdb/query/v4 calls fail with a 500.
- Add puppetdb_read role via CNPG managed.roles with password sourced
from a new postgres-read-credentials Vault secret
- Grant CONNECT, USAGE, SELECT and default privileges to puppetdb_read
via postInitApplicationSQL (must also be run manually on existing cluster)
- Add puppet-postgres-pooler-ro Pooler (type: ro) routing to replicas
- Add puppetdb-read-database-conf ConfigMap with read-database.conf
mounted into /etc/puppetlabs/puppetdb/conf.d/ in the PuppetDB deployment
- Wire OPENVOXDB_READ_POSTGRES_* env vars from the new secret
💘 Generated with Crush
Assisted-by: Claude Sonnet 4.6 via Crush <crush@charm.land>
Reviewed-on: #69
- master config section is not used
- server containes all setting specifically for a server (puppet, puppet ca)
- user is for all puppet <command> tooling, like 'puppet generate'
Reviewed-on: #66
Add support for installing additional Ruby gems via custom entrypoint script.
The script is mounted as a ConfigMap into /container-custom-entrypoint.d/
and will be executed during Puppetserver container startup.
Reviewed-on: #63
- Mount vault-ca-cert secret at /opt/vault-ca-cert.crt in both deployments
- Update cobbler-enc script to use correct CA certificate path
- Resolves OSError about missing TLS CA certificate bundle
Reviewed-on: #62
- Add puppet-shared-bins PVC (10GB) for shared binaries
- Mount /opt/bin in both compiler and master deployments
- Add init container to install uv binary and cobbler script to shared volume
- Update cobbler-enc to use absolute path and uv cache directory
- Configure puppet.conf to reference cobbler-enc from /opt/bin
Reviewed-on: #61
DEBUG:urllib3.connectionpool:Starting new HTTP connection (1): puppetdb:8081
ERROR:pypuppetdb.api.base:Could not reach PuppetDB on puppetdb:8081 over HTTP.
- puppetdb_host assumes HTTP when not verifying ssl
Reviewed-on: #58
- Update PuppetDB connections from HTTP (8080) to HTTPS (8081)
- Add automatic certificate generation for Puppetboard using Puppet CA
- Implement initContainers for proper certificate provisioning before app start
- Add dedicated PVC for Puppetboard certificates with RWX access
- Configure SSL verification and client authentication for secure PuppetDB access
Reviewed-on: #50
This resolves SSL certificate verification failures preventing puppetdb access
- Update OPENVOXDB_SERVER_URLS from https://puppetdb:8081 to http://puppetdb:8080
- External access to puppetdb will still use HTTPS via ingress
- Internal cluster communication does not require encryption
Reviewed-on: #47
- change puppet/puppetca -> LoadBalancer
- dedicate ip's for puppet and puppetca loadbalancers
- name the puppetserver port
- remove puppet/puppetca ingress
Reviewed-on: #35
puppetdb_port has tcp:// in it, even though we pass the correct variable
in from a configmap.
```
ben@metabox ~/s/p/argocd-apps> kubectl --context admin run debug-pod --image=busybox --rm -it --restart=Never -n puppet -- env | grep -i puppetdb_port
PUPPETDB_PORT_8081_TCP_PORT=8081
PUPPETDB_PORT_8081_TCP_PROTO=tcp
PUPPETDB_PORT=tcp://10.43.101.142:8080
PUPPETDB_PORT_8080_TCP=tcp://10.43.101.142:8080
PUPPETDB_PORT_8080_TCP_ADDR=10.43.101.142
PUPPETDB_PORT_8081_TCP=tcp://10.43.101.142:8081
PUPPETDB_PORT_8080_TCP_PROTO=tcp
PUPPETDB_PORT_8081_TCP_ADDR=10.43.101.142
PUPPETDB_PORT_8080_TCP_PORT=8080
```
Reviewed-on: #32
the puppetca is not pointing to the puppetmasters which prevents the
puppet-compilers from starting, preventing puppetdb/puppetboard from
starting.
- point puppetca service -> puppetserver-master
Reviewed-on: #31
updating all the names of services and their respective filenames to
better match the way puppet infra is used in my lab.
- puppet -> the compilers
- puppetca -> the master(s)
- puppetdb -> the puppetdb
- puppetboard -> puppetboard
updated references to these services in all other definitions I could find
note: need a good way to test these changes with argocd
Reviewed-on: #30
complete the implementation of puppet in kubernetes, taking many
features from the openvox helm chart and improving on them. changes from
helm are:
- using vault for storing secrets
- using g10k instead of r10k
- using a single shared g10k cronjob for all masters/compilers
- using a single shared /etc/puppetlabs/code directory (shared, cephfs)
changes:
- deploy puppet master and compiler servers with statefulset/deployment
- deploy puppetdb with postgresql backend, taking advantage of cnpg cluster and pooler
- deploy puppetboard
- all supporting configmaps, services, ingresses, and hpas
- added vaultstaticsecret for eyaml private keys
- configured secure mounting of eyaml keys at /var/lib/puppet/keys/
- updated base kustomization to include all 23 new puppet resource files
Reviewed-on: #29
g10k hardlinks, so reqires that the cache and code be in the same pvc.
updated r10k repository with cachedir in same pvc, and so now I can
remove these unused pvcs from argo.
unkin/puppet-r10k#4
Reviewed-on: #28
unkinben
benvin