*** THE BUG, AND IT WAS MINE. *** The placeholder guards were `case "$s3" in
""|TBD-*)` — a PREFIX glob. Two commits earlier (32d6766) I changed the
placeholder from `TBD-PROD-S3-BACKUP-ENDPOINT` to
`https://TBD-R2-ACCOUNT-ID.r2.cloudflarestorage.com`, which does NOT start with
`TBD-`. Verified: fnmatch('https://TBD-R2-...', 'TBD-*') is False. All three
guards silently stopped matching, so **step_resume would have resumed velero on
prod straight onto a nonexistent hostname** — velero installs, reports healthy,
and no backup exists. That is precisely the failure the guard was written to
prevent, disarmed by a value change in a DIFFERENT file with nothing tying the
two together. Replaced with substring-matching s3_is_placeholder in preflight,
s3check and resume.
s3check rewritten from a reachability ping into a real proof. Per credential, from
a throwaway in-cluster pod (pod egress is the only egress that matters — the
workstation sits on a different network side and a host-side reading has already
misled this project once):
1. the credential LOADS AND AUTHENTICATES in the exact shape its consumer reads
— Velero's INI mounted from the `cloud` key, CNPG's discrete env keys;
2. the bucket EXISTS and is LISTABLE under this environment's own prefix, via
ListObjectsV2 — the op Velero's BSL sync and Barman's WAL scan actually use.
ListBuckets deliberately avoided: a bucket-scoped R2 token may not grant it;
3. PUT -> GET -> sha256+length compare -> DELETE -> confirm-gone. A HEAD would
pass with a read-only token and with every write being rejected;
4. the R2 checksum trap probed ON PURPOSE by repeating the PUT with checksums
enabled, and named either way.
Buckets and prefixes are DERIVED (velero's from the authored HelmRelease via yq,
cnpg's parsed from CNPG_BACKUP_PATH), not restated.
Fails closed, with no "unknown, carrying on" arm left: no pod, unfinished pod,
empty log, missing/duplicate/malformed verdict, any failed assertion, a claimed
PASS with failures>0, an assertion count below the caller-computed minimum, an
OK-line count disagreeing with the claimed count, or a PASS from a pod that ended
Failed. Failures are classified by cause — no egress, NoSuchBucket,
InvalidAccessKeyId, AccessDenied (token not scoped to BOTH buckets),
InvalidArgument -> the checksum trap.
`amazon/aws-cli` pinned rather than `minio/mc`: mc lacks sed/awk (which already
broke one probe) AND cannot send checksum headers, so it cannot test for the trap.
TIER_ORDER now includes platform-config-cdi, between platform-config and the leaf
tiers. Without it `resume` returned success without ever waiting on the tier that
guarantees the CDI StorageProfiles — and apps-vms depends on them: unifi
provisions on longhorn-vm, where an unapplied profile means CDI infers RWO instead
of the authored RWX, permanently non-live-migratable.
CREDENTIALS: one key, four committed Secrets, two shapes — and NO manifest change
is needed, only encrypted values. platform/backup/set-r2-credentials.sh writes all
four with `sops set`, which encrypts IN PLACE, so no decrypted Secret ever exists
in the working tree to be committed by accident — strictly better than staging a
plaintext template. Prompts with `read -rs`, rejects anything outside the S3
credential charset (a paste that swallowed a newline authenticates nowhere), never
prints a value, and verifies by decrypting back and comparing fingerprints.
`--check` is read-only. The MinIO key is RETAINED — it is what reads the
pre-migration backups.
THE ACCOUNT ID IS ALREADY IN THE REPO: infra/tofu/backend.r2.hcl commits
https://7ce452a1b3d6c86daf2c020edcb82d37.r2.cloudflarestorage.com for the tofu
state bucket. If the backup buckets go in the same account, only the two token
values are needed. The placeholder is KEPT anyway — fail-closed beats convenience,
and the guard above just demonstrated why.
Corroborating the checksum fix independently: infra/tofu/backend.tf already carries
`region = "auto"` and `skip_s3_checksum = true # R2 rejects AWS trailing
(streaming) checksums`. This repo had learned both R2 facts for tofu state already.
VERSION FINDINGS, documented not measured against R2: plugin v1.14.2 documents
checksumAlgorithm defaulting to CRC32, so the authored "" is load-bearing;
Velero's kopia backend has NO checksum knob, so BSL metadata and PVC-data
compatibility are SEPARATE questions; and in the live CNPG 16.4 image,
boto3/botocore 1.35.60 is below the 1.36 line that turned on default request
checksums — so CNPG should be clear today, but barman-cloud exposes no escape
hatch, making a CNPG image bump an R2 re-verification event.
VERIFIED BY ME, independently of the agent: the dead-guard glob (fnmatch False);
the placeholder now refused on prod before any pod is created; a full round trip
against dev's MinIO — 7/7 assertions on BOTH credential paths, byte-identical,
pod Succeeded, nothing left behind, checksum trap correctly reported absent; all
four Secrets and the committed account id present. Full CI gate PASS.
STILL UNPROVEN AND SAID SO IN THE DOC: Velero-to-R2 (metadata or kopia data),
CNPG-barman-to-R2, and any restore from R2. R-STATE-03 is PROVEN MinIO-ONLY and
its ledger row predates this move. infra/BACKUP-R2.md §5 is the re-proof sequence
and it ends in a RESTORE, not an upload.
Refs R-STATE-03, R-IDP-05, R-ENV-02.
Co-Authored-By: Claude Opus 5 <[email protected]>
`platform-config` applied the CDI CR *and* cdi-storageprofiles.yaml in ONE pass.
StorageProfile's CRD does not exist until cdi-operator reconciles that CR, and
Flux server-side dry-runs every object before applying any — so the unknown kind
aborted the tier, the CR was never applied, the CRD was never created, and the
tier could NEVER converge. Observed on prod: cdi-operator Available 12h, no CDI
CR, no storageprofiles CRD, every downstream tier blocked.
The profiles now live in their own tier ./platform-config-cdi with
dependsOn: platform-config and wait: true. Because dependsOn is REVISION-gated
and platform-config runs wait: true, the CDI CR must be applied AND healthy —
exactly the event that creates the CRD — before the profiles are applied. Cold
clusters order themselves. `standup.sh crds` is retained as a fallback for
clusters already stuck, and because it is idempotent.
THE GENERALISABLE RULE (recorded in clusters/README.md): a CR whose CRD is
created by a CONTROLLER AT RUNTIME is second-order and needs its own pass.
cdi-operator.yaml registers exactly ONE CRD (cdis.cdi.kubevirt.io); storageprofiles,
datavolumes, datasources and dataimportcrons appear only after it reconciles the
CDI CR. An audit found StorageProfile is the only second-order kind in
platform-config — KubeVirt, CDI, KeycloakRealmImport, CNPG Cluster,
VolumeSnapshotClass, ClusterIssuer, PrometheusRule, ServiceMonitor and the Cilium
L2 CRs all have their CRD in git or in a chart.
Moving the CDI CR up into `platform` was rejected: that would put CDI's health on
the tier the CNI gates and everything waits on — the mistake the oauth2-proxy and
headlamp splits exist to undo.
apps-vms gains dependsOn: platform-config-cdi. LOAD-BEARING, not cosmetic:
apps/vms/unifi/20-datavolume.yaml provisions on storageClassName: longhorn-vm and
its comment claimed platform-config guaranteed the profiles — which this move
would have silently falsified.
THE PRUNE TRANSIENT, MEASURED RATHER THAN ASSUMED. platform-config has
prune: true, so it deletes both profiles once, CDI recreates them with empty spec,
and the new tier re-applies the overrides ~30-60s later. Compared our override
against CDI's unaided inference for the same Longhorn provisioner on dev:
longhorn-vm (we override) spec RWX/Block -> status ReadWriteMany/Block
longhorn-static (no override) spec unset -> status ReadWriteOnce/Block
Both infer cloneStrategy: snapshot. So cloneStrategy is NOT at risk; accessModes
IS — unaided CDI gives RWO. A DataVolume provisioned on longhorn-vm inside the
window would be stamped RWO/Block and be PERMANENTLY non-live-migratable. Flux can
no longer cause it (apps-vms gates on the new tier); a human must not do it by hand.
Corroborating evidence that the recreate happens at all: dev has 5 StorageClasses
and 5 StorageProfiles, three with fluxOwner=NONE, spec unset and status populated
— CDI creating and inferring them with no Flux involvement.
MERGE PRECONDITION: `flux get kustomization platform-config` must be Ready on the
target first. If it is not, the profiles are pruned and sit at CDI defaults until
it recovers — and prod was in that state hours earlier. Fallback is the two-step
the Cilium move used (merge with prune: false, verify the ownership label moved,
restore prune: true).
substituteFrom: cluster-vars is declared even though the manifest has no ${TOKEN},
deliberately: conformance/tests/l0-static/flux-postbuild-substitution only audits
paths whose Kustomization declares it, so omitting it would move the file OUT of
the audited set — a blind spot for a future literal-token ship. Gate confirms it
stayed in (path=./platform-config-cdi files=2 defined=0 hazards=0). No SOPS
decryption stanza: the moved file is two plain CRs, so no overlay change was needed.
VERIFIED INDEPENDENTLY OF THE AGENT: platform-config renders 0 StorageProfile
while keeping the CDI and KubeVirt CRs; platform-config-cdi renders exactly the
two profiles; the graph is byte-identical across dev and prod with apps-vms
carrying both edges and no cycle. Full CI gate PASS.
KNOWN FOLLOW-UP: bootstrap/standup.sh TIER_ORDER does not yet include
platform-config-cdi, so `resume` will not wait on it. Another agent holds that
file; the PM will add it.
Refs R-GITOPS-01, R-ENV-02.
Co-Authored-By: Claude Opus 5 <[email protected]>
"That's where backups will go for the time being until a better on-prem backup
solution is created." Two reasons beyond the (negligible) cost:
1. Prod CANNOT reach the on-Proxmox MinIO. Prod is on the SOHO network
192.168.2.0/24 — pfSense's WAN — while that MinIO sits inside pfSense's LAN.
Measured from a pod inside prod: connection timed out after 15s. R2 needs no
firewall change.
2. That MinIO runs on the SAME Proxmox host as dev's VMs, so dev's
"off-cluster" backup shared a physical failure domain with dev — one host
loss took the cluster AND its backups. R2 is genuinely off-site.
COST, measured rather than estimated (read straight out of the buckets):
velero 4.0 GiB / 6,264 objects (7d retention; 8.65 GB live -> kopia dedup)
cnpg 211 MiB / 1,755 objects (keycloak 66Mi/331, nextcloud 87Mi/1329,
semaphore 58Mi/95; ~7d of a 30d window)
Both clusters project to ~13-15 GiB and ~30k writes/month => pennies on R2, and
free egress means restore REHEARSALS cost nothing, so R-STATE-03 has no financial
disincentive to stay proven. Wasabi/Hetzner are 50-200x worse here purely because
of 1 TB minimums; Glacier tiers are a trap (90-day minimum billing against a
7-day Velero retention plus early-delete fees on 30-day WAL pruning).
*** THE R2 GOTCHA, WIRED UP FRONT: checksumAlgorithm: "" ***
velero-plugin-for-aws v1.14.2 bundles an AWS SDK that sends
x-amz-sdk-checksum-algorithm / x-amz-checksum-crc32 on every PUT. R2 REJECTS
those with HTTP 400 InvalidArgument, so every backup and every kopia blob upload
would fail. Set unconditionally — MinIO ignores it, so it needs no fork.
`region` was hardcoded us-east-1 inside the Velero HelmRelease: correct-by-
accident for MinIO (which ignores it) and WRONG for R2 (which demands exactly
`auto`). Now the S3_BACKUP_REGION knob.
PREFIXES ARE NOW SYMMETRIC (dev/prod), reversing the reasoning from two commits
ago. While the target was MinIO, dev HAD to keep the bucket root or it would
orphan the backups its proven restore depends on. R2 starts BOTH clusters on an
empty bucket, so there is no history to orphan — and leaving the bucket root
unused means a misrendered prefix lands somewhere obviously wrong instead of
silently colliding with a real environment.
NEITHER OVERLAY PINS THE ENDPOINT ANY MORE — a deliberate reversal of this
repo's "prod must never inherit base" rule. That rule exists because base
defaults are dev's LAN addresses; R2 is a single environment-INDEPENDENT endpoint,
so under R-ENV-02 it belongs in base, and two copies of one URL is a rotation
waiting to drift. Isolation comes from the path knobs, not the URL. Caught while
verifying: dev's overlay DID still pin MinIO, so dev would have taken the new
`dev` prefix while writing to the old endpoint — orphaning its MinIO backups for
no benefit. That is why the endpoint and the prefixes must move in ONE commit.
*** MERGING THIS RESETS DEV'S PITR WINDOW. *** Dev's R2 prefix starts empty:
Velero writes a fresh kopia repo, CNPG takes new base backups, the 30-day Barman
window rebuilds from zero. The MinIO copy is NOT deleted and stays restorable by
pointing the endpoint back. Do not merge while dev's proven restore is the only
recovery path you rely on.
STILL BLOCKED ON THE OPERATOR: the endpoint is the loud placeholder
https://TBD-R2-ACCOUNT-ID.r2.cloudflarestorage.com. Creating the R2 buckets and
an API token needs Cloudflare account access I do not have (the token in this
repo is Zone:DNS scoped). Velero+R2 and CNPG-barman+R2 are also UNPROVEN here and
must be verified by an actual backup AND restore, not just an upload.
VERIFIED: full CI gate PASS. Rendered both overlays — dev dev/s3://cnpg/dev,
prod prod/s3://cnpg/prod, region auto, one shared endpoint.
Refs R-STATE-03, R-ENV-02, R-IDP-05.
Co-Authored-By: Claude Opus 5 <[email protected]>
Operator ruling: "the S3 backup location is the same as Dev." Prod's
S3_BACKUP_ENDPOINT is now http://192.168.1.9:9000 and the TBD-PROD-* placeholder
is retired. Set EXPLICITLY rather than by omitting the op — an omitted op would
inherit the same value silently, and a shared backup target must be visible in
the overlay that owns it.
SHARING THE ENDPOINT WAS NOT SAFE AS AUTHORED. Nothing about the backup PATHS
was per-environment, and enabling prod against dev's MinIO would have been the
2026-09-04 external-dns incident again, in a worse place:
VELERO — bucket `velero`, NO prefix, and both clusters run schedules named
`daily` / `daily-resources-only`. Velero SYNCS backups it finds in its BSL
into itself as Backup CRs and GARBAGE-COLLECTS them at TTL, so prod would
have adopted dev's backups and then deleted them on schedule.
CNPG — the worse one. destinationPath `s3://cnpg` with CNPG appending
serverName, which defaults to the CLUSTER NAME. Both environments have
keycloak-db, nextcloud-db and semaphore-db, so prod's Postgres would archive
WAL into the EXACT prefix dev's does. Two servers interleaving WAL into one
Barman stream CORRUPTS THE CHAIN and destroys PITR — and dev's restore is
PROVEN (R-STATE-03). Sharing an endpoint is fine; sharing a path is not.
Two knobs, authored once, environments differing only by scoped overrides
(R-ENV-02):
VELERO_BACKUP_PREFIX dev "" | prod "prod"
CNPG_BACKUP_PATH dev "s3://cnpg" | prod "s3://cnpg/prod"
THE BASE VALUES ARE DEV'S CURRENT LAYOUT ON PURPOSE. Dev's existing backups live
at the bucket root and at s3://cnpg; giving dev a prefix would ORPHAN the very
backups its proven restore depends on — the same reasoning that keeps dev's
EXTERNALDNS_TXT_OWNER_ID at `default`. Prod has no history to preserve.
*** PROD STILL HAS NO BACKUP COVERAGE — the endpoint is not reachable. ***
Measured from a throwaway pod INSIDE the prod cluster (the workstation proves
nothing about pod egress, and a workstation-side reading has already misled this
project once): curl http://192.168.1.9:9000/minio/health/live → connection timed
out after 15s. pfSense does not route prod's pod network to the dev LAN's MinIO.
OPERATOR ACTION REQUIRED. velero stays suspended; note apps-wave-0 applies GREEN
regardless — the token IS defined so strict substitution succeeds — while every
write fails at RUNTIME. A green apps-wave-0 does not mean backups work.
New `standup.sh s3check` step probes the endpoint from a throwaway in-cluster
pod and fails closed, with `--s3-endpoint <url>` to test a CANDIDATE before
authoring it into cluster-vars. That is how the unreachability above was
established rather than assumed.
VERIFIED: full CI gate PASS. flux-postbuild-substitution 254 -> 260 assertions,
and neither new knob appears in the dead-knob advisory list, so both are really
consumed. Rendered both overlays: all three CNPG destinationPaths and Velero's
prefix carry the tokens; dev resolves to its existing layout, prod to the nested
paths.
DELIVERY: prod reconciles the bootstrap Gitea's main at c8c1d29. This branch also
carries the DNS inversion, which must NOT reach prod. Delivering needs a
c8c1d29-based branch with only these commits — no urgency while the routing is
still closed.
Refs R-STATE-03, R-ENV-02, R-IDP-05.
Co-Authored-By: Claude Opus 5 <[email protected]>
*** DO NOT MERGE TO main WHILE THE OPERATOR IS AWAY. MERGING IS THE MIGRATION. ***
Operator simplified the architecture, inverting c8c1d29:
dev -> *.dev.home.keithjackson.dev (was the apex)
prod -> *.home.keithjackson.dev (the apex; no prod. subdomain, no CNAME layer)
WHY MERGING IS DANGEROUS, and this is the finding that matters most here:
clusters/dev/sync.yaml self-reconciles ./clusters/dev from `main` every 10 minutes. Merging
these files EXECUTES the cutover on the live dev cluster, unattended. And because the same
commit moves dev's GitRepository url to gitea.dev.home.keithjackson.dev, dev STRANDS — it
stops reconciling and cannot be fixed by pushing, because the broken thing IS the path Flux
reads pushes through. Banners added at the top of clusters/dev/kustomization.yaml and beside
the url itself. That url cannot be tokenized: the root Kustomization declares no
substituteFrom, so `gitea.${BASE_DOMAIN}` would ship the literal token.
Verified dev is currently safe: this work is NOT on origin/main, dev is reconciling
main@4809199, and its live BASE_DOMAIN is still home.keithjackson.dev. Deliberately NOT
pushed to the bootstrap Gitea either — prod reconciles from there, and prod on `prod.home.`
cannot collide with dev on the apex, so that is the safer interim for prod.
VALUES
BASE_DOMAIN TXT_OWNER_ID DOMAIN_FILTER
base dev.home.keithjackson.dev sentinel keithjackson.dev
dev dev.home.keithjackson.dev dev keithjackson.dev
prod home.keithjackson.dev prod keithjackson.dev
Ansible seam inverted too: role default k3s_base_domain = dev's subdomain, group_vars/k3s_prod
overrides to the apex.
dev's txtOwnerId moves `default` -> `dev`. The old value was FORCED (its live apex records
were already registered under `default`, so renaming would have orphaned them). On a new
suffix dev publishes a brand-new record set with no prior registry, so the id is chosen on
merits — and `default` has none: it is what you get by FORGETTING the flag, which is exactly
how both clusters came to share it and delete each other's records. Operationally it also
means dev's old apex records stay owned by a now-unclaimed id, so nothing deletes them
implicitly and vacating the apex becomes an explicit sequenceable step rather than a race
between two reconcile loops. PROD MUST NEVER BE GIVEN `default`: prod inherits the apex
NAMES, so `default` is the one value that would let it adopt and policy:sync-delete dev's
still-live apex records — the same incident with the roles swapped.
DOMAIN_FILTER stays the ZONE for both, with F-DNS1 written into all three files as a
do-not-retry: a subdomain filter drops the parent zone and publishes NOTHING.
VERIFICATION — the property changed, and so did the method. "Dev renders byte-identically" no
longer applies because dev is deliberately migrating. Instead: apply the INVERSE substitution
to the new render and demand byte-identity with the baseline. 145 changed line pairs, every
one classified: 143 hostname-move, 2 declared owner-id, ZERO unexplained, ZERO bare-apex
remnants. cilium and velero show 0 changes and provably contain no hostname, so they are not
blind spots. Prod: 0 occurrences of prod.home., 0 of dev.home., 143 apex names. Certs: dev
*.dev.home.keithjackson.dev, prod *.home.keithjackson.dev.
THREE positive controls, all fire — including one that proves the two checks are not
redundant: injecting a single un-moved hostname SURVIVES the inverse-substitution check and is
caught only by the independent bare-apex scan. That is why both exist.
Ansible probe (connection: local, zero SSH): dev nodes -> keycloak.dev.home..., prod nodes ->
keycloak.home.... Used roles: + --tags probe rather than include_vars, because include_vars
outranks group_vars and would have FALSELY shown prod inheriting dev's value.
Gates all PASS: l0, l1, render-golden (10 goldens), lint, substitution (254 assertions),
render.py --selftest, spec-immutability. Catalog ServiceDefinitions edited, never rendered apps/.
RUNBOOK rewritten in §10 + §10a with the ordering that avoids stranding: D0 snapshot escape
routes -> D1 break the automatic trigger -> D2 publish and PROVE gitea.dev.home + the new
wildcard cert WHILE dev still serves the apex -> D3 merge -> D4 release reconciliation
deliberately -> D5 Keycloak issuer + ~12 clients (headlamp's is SOPS, by hand) -> D6
one-server-at-a-time k3s restart + SSO re-issue (service restart only, NO reboots) -> D7
retire old apex records -> D8 verify the apex is empty AGAINST CLOUDFLARE, not pfSense —
the gate on everything after. Then P1 unsuspend prod.
ACME: prod's wildcard is EXACTLY the name set dev holds today, so prod's orders share dev's
5-per-week duplicate-certificate bucket. Order after D8, delete dev's old Certificate rather
than let it renew, rehearse on staging.
The 6 L2 chainsaw tests hardcoding the apex are enumerated but deliberately NOT edited — while
dev is live on the apex they are correct; they need per-environment parameterisation.
Serves: R-ENV-02, R-DNS-01/02/03, R-TLS-01.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
INCIDENT FIX. Prod's external-dns DELETED EVERY ONE OF DEV'S PUBLIC DNS RECORDS. Both
clusters ran policy: sync against the SAME Cloudflare zone with the SAME implicit
txtOwnerId (`default`), so prod believed it owned dev's records, found no matching sources,
and removed them plus their TXT registry entries. Dev recovered ~30s after prod was scaled
to 0; verified restored against Cloudflare directly (my first check went through a pfSense
override and falsely showed healthy — local resolvers are not evidence about public DNS).
THE MISS WAS MINE, and it was a failure of imagination in how I briefed every agent. I put
"never touch dev" in brief after brief and enforced it — separate kubeconfigs, verified
--limit, sha-checked dev's kubeconfig before and after. Every one of those guards held. NONE
of them covered SHARED EXTERNAL RESOURCES. Prod reconciling its OWN correctly-authored
manifests reached across a third party and destroyed dev's state. Cluster-to-cluster
isolation was verified; the dependency that mattered was cluster-to-Cloudflare.
Operator's architecture: dev -> *.dev.home.keithjackson.dev, prod ->
*.prod.home.keithjackson.dev, apex CNAME <svc>.home -> <svc>.prod.home.
PHASE 1 ONLY (this commit): prod moves to its own subdomain. DEV IS NOT MIGRATED and its
effective hostname stays home.keithjackson.dev, so dev renders byte-identically and nothing
live moves. That alone ends the collision and unblocks prod. Phase 2 (dev -> dev.home. +
apex CNAMEs) is deferred to when the operator is present: it relocates Keycloak's issuer and
12 clients' redirect URIs, forces a rolling k3s restart and kubectl SSO re-issue for every
user, and moves gitea.home. — which is BOTH the GitOps source and the current push target.
BASE_DOMAIN WAS A DEAD KNOB — it existed in cluster-vars but the status report flagged it as
referenced by nothing, because hostnames were hardcoded throughout. Now real: 141 refs.
Searched the AUTHORITATIVE reconciled set by resolving each Flux spec.path transitively via
conformance/lib/status.py — the same resolver the L0 gate uses, so "what Flux reconciles"
has one definition. 185 occurrences across 47 files in 154 reconciled files. Tokenized:
Gateway listeners, the wildcard Certificate, coredns-custom, external-dns, Keycloak
instance/realm (26 redirect URIs), oauth2-proxy, Grafana OIDC, all HTTPRoutes, Homepage's 15
links, 30 runbook_urls, unifi cloud-init, render.py's ZONE constant (which drove
--oidc-issuer-url/--cookie-domain/--whitelist-domain on every forward-auth sidecar) and 11
catalog/ ServiceDefinitions re-rendered canonically. Left alone: 43 comment/prose
occurrences and non-reconciled paths.
TWO THAT CANNOT BE TOKENIZED, flagged not forced:
- headlamp's OIDC URLs live inside a SOPS-encrypted Secret; re-encrypting rewrites the MAC
over every value and the OFFLINE harness cannot verify it (kustomize does not decrypt),
so an error would break live-dev SSO with no pre-flight proof. Prod needs its own copy
anyway — client secrets are realm-scoped.
- kube-apiserver --oidc-issuer-url is a k3s SERVER FLAG, not a manifest, so Flux can never
reach it. Done on the inventory seam instead (k3s_base_domain -> k3s_oidc_issuer_url,
overridden in group_vars/k3s_prod). Verified per host: dev keeps the default, prod gets
prod.home.
EXTERNAL-DNS OWNER, the actual incident cause: txtOwnerId is now explicit per environment —
dev renders exactly `default` (the chart default it was already using, so its existing
records stay recognised) and prod gets `prod`. The BASE default is a NON-COLLIDING SENTINEL,
deliberately not `default`, so a future environment that forgets the knob owns nothing and
can delete nothing rather than silently inheriting dev's identity.
I OVERRODE ONE AGENT DECISION against our own prior finding: it narrowed prod's
domainFilters to prod.home.keithjackson.dev as defence in depth. FINDING F-DNS1
(DECISIONS-LOG 2026-08-17) already proved a subdomain filter DROPS THE PARENT ZONE —
ExternalDNS generates the record then skips it with "no hosted zone matching record DNS
Name". Prod would have published NOTHING. Set to the ZONE (keithjackson.dev) like dev.
Isolation comes from the subdomain split plus txtOwnerId, which is what the incident
actually needed and what a domain filter cannot provide.
DEV PROVEN UNCHANGED: 11 of 12 substituting paths byte-identical; the 12th differs by
exactly one line, `+ txtOwnerId: default` — the value dev was already using implicitly.
POSITIVE CONTROLS confirm the harness is not blind: perturbing BASE_DOMAIN moves 10 of 12
hashes (the 2 that don't provably contain no hostname), perturbing the owner id moves
exactly 1. Prod's rendered output has ZERO home.keithjackson.dev references outside prod.,
and its cert renders *.prod.home.keithjackson.dev — no overlap with dev's.
L0 substitution assertions 106 -> 254; BASE_DOMAIN is no longer flagged dead. Gate PASS.
PROD REMAINS SUSPENDED (platform + 10 downstream tiers) pending review; nothing applied.
Serves: R-ENV-02, R-DNS-01, R-DNS-02, R-DNS-03, R-TLS-01.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Prod is being stood up now, without waiting for node 2's RAM upgrade. Three coupled
changes, all applied and verified on all three nodes.
=== 1. KERNEL >= 6.13 — now a PREREQUISITE, not a nice-to-have ===
A k3s SERVER that intermittently boots with 2.6 GiB is not a degraded node; it fails under
etcd plus its workload share, unpredictably, on any reboot. Fixed before k3s goes on.
roles/node_baseline gains a `kernel`-tagged block: asserts the suite matches the node's
Debian release, writes a deb822 trixie-backports source, installs, reads back the exact
version, writes a provenance/rollback note to the node. Scoped by
node_baseline_backports_kernel (default FALSE, true only in group_vars/k3s_prod) — chosen
over a hostname test so a fourth prod node inherits it from group membership. Verified:
--check against dev skips all 8 tasks on all 3 hosts. Two deliberate calls: `state: latest`
not `present` (linux-image-amd64 was already installed from main, so `present` would have
silently fixed nothing), and NO preferences.d pin (backports' NotAutomatic +
ButAutomaticUpgrades already gives 100/500; an explicit Pin-Priority: 100 would switch OFF
automatic security upgrades for the very kernel being installed).
Installed 7.1.8+deb13-amd64 on all three; 6.12.107 kept bootable.
THE DECISIVE PROOF: node 2 booted at MEMATTR=0xbeeb0018 — the EXACT address that gave it
2.6 GiB twice — and now reserves 634,904K instead of 13,958,980K. MemTotal 15.4 GiB. Same
address, handled correctly; not a lucky draw.
§0b's claim that a fixed kernel prints 'FW_BUG Corrupted...' is CONDITIONAL and was
corrected: >=6.13 has three rejection gates and only the third prints that line; all three
nodes hit gate 1 (version > 2). Proof instead is boot-independent — decompressing
/boot/vmlinuz on all three yields the FW_BUG and descriptor-size strings verbatim; neither
exists in 6.12.
TWO MORE BUGS IN MY OWN HEALTH CHECK, found because the fix couldn't otherwise be shown:
- MEMATTR_VERSION_RE matched (\d+) but the kernel prints that field with %d — SIGNED. Node
2's real value was -1569980416, which did not match, so state fell through to `clean` and
the node reported PASS "the kernel accepted it without complaint". A FALSE PASS on the
exact bug the check exists to find. Now (-?\d+).
- bad_version was WARN unconditionally, so a FIXED node would have stayed WARN forever.
Now keys on memattr_kernel_fixed: same line, WARN below 6.13, PASS at/above.
=== 2. DISK LAYOUT — Longhorn claimed, per-node paths DISCOVERED not hardcoded ===
Operator decision: reuse the ~424 GB partition; accept the shared-device I/O risk (dev's
etcd/Longhorn contention was on a spinning HDD, prod is SSD/NVMe).
Device paths differ per node (sda5 / nvme0n1p5 / sda5), so the role DISCOVERS the partition
via `findmnt --mountpoint` on a declared claim mountpoint rather than trusting inventory.
Per-host host_vars would have frozen today's hardware into inventory, where it goes stale on
the first re-image and the stale value's failure mode is mkfs on the WRONG partition.
Eleven fail-closed asserts before any write: real block device, not backing / /boot /usr
/var /home /tmp or active swap (built from the node's OWN gathered facts, never a hardcoded
list), already ext4, no foreign label, >= 100 GiB, and empty. Relabel, never reformat.
/srv verified empty first on every node (ls -A -> lost+found only; du 20K; the "5.1% used"
was ext4's default 5% root reserve).
=== 3. k3s DATA-DIR MOVED — /var was 11.8 GiB and would have killed etcd ===
Measured on the fully-deployed dev cluster: containerd 21-25 GB/node, /var total 58-70 GB.
Prod's /var is 11.8 GiB and held BOTH etcd and the image store, so prod would have filled
/var during its first image pulls — and a full /var means etcd cannot write its WAL. Cluster
death, not degradation, arriving during bring-up. Invisible when the disk decision was made,
because "etcd on root" sounded like the 30 GiB / rather than a separate 12 GiB /var.
Now: p5 -> /data (LABEL=k3s-data), with /data/longhorn and /data/k3s. Neutral names chosen
deliberately — the partition now has two tenants, and reusing /srv would make "claimed" and
"unclaimed" indistinguishable to the idempotency pivot, while a bare `data` label would make
a stranger's `data` volume look already-claimed and weaken an assert.
The transition off the existing LABEL=longhorn mount is DECLARED, not inferred
(node_baseline_data_retired_fs_label/_retired_mount, defaults ""), because changing the label
makes a claimed node answer "no" to the pivot and walk into the greenfield path. Self-retiring
once blkid -L k3s-data succeeds. THREE independent gates keep live dev out: retired vars
default ""; gated on longhorn_source == claim-mountpoint (dev is `device`); and the emptiness
assert fails closed on dev's 18 volumes even if both were defeated.
k3s_data_dir is one variable read by two roles (k3s_server --data-dir; node_baseline derives
the etcd mount). Role default is k3s's own default, so dev is unchanged and keeps its
dedicated etcd device. node_baseline_etcd_source becomes `shared-data-partition` and is
LOAD-BEARING: the role asserts per run that the filesystem backing the etcd dir really IS the
data partition — had the mount silently failed, etcd plus 21-25 GB of images would land back
on the OS filesystem invisibly until the disk filled.
LONGHORN_DATA_PATH is a new cluster-vars knob (R-ENV-02): base and DEV /var/lib/longhorn,
prod /data/longhorn.
DEV PROVEN UNCHANGED, which is the property that mattered — the live cluster has 18 Longhorn
volumes on that path. All 12 substituting Flux paths rendered through
`kubectl kustomize | flux envsubst --strict` for both trees: 12/12 per-file SHA-256 identical
and an identical aggregate SHA over every rendered dev byte. With a POSITIVE CONTROL showing
the same harness detects prod changing on exactly one line — an all-identical result is
worthless if the harness cannot see change. Live check after: dev's default-data-path setting
still /var/lib/longhorn, 18 volumes attached, 0 Kustomizations unready.
Idempotent: run 2 is changed=0 on all three nodes. Also fixed a real tag bug — include_tasks
does NOT propagate its own tags to included tasks (unlike import_tasks), so `--tags disks`
selected the include and ran nothing inside it while reporting success.
OPEN DEFECTS, tracked not forgotten: node 2's warm-reboot hang (one clean reboot since, NOT a
fix); and node 3 runs dhcpcd for its wifi interface, which rewrites /etc/resolv.conf on lease
renewal — the role's `contended` detection checks resolvconf and NetworkManager only, so node
3's decided resolvers will silently revert.
Gate: MERGE GATE: PASS. L0 substitution check 106 assertions.
Serves: R-ENV-01, R-ENV-02, R-ENV-03, R-STATE-01, R-STATE-02, R-PLAT-01. Task: T-7.1.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Cilium was one of ~12 components inside a single `platform` Kustomization with wait:true,
so the CNI's health was indistinguishable from eleven unrelated components in the tier's
Ready condition, and the only emergency valve was freezing the whole platform tier. This
repo already set the precedent in reverse (oauth2-proxy and headlamp were split OUT so
unhealthy leaves would not gate a tier); Cilium is the mirror image — the one component
that legitimately SHOULD gate everything.
New clusters/base/cilium.yaml: path ./platform/cilium, wait:true, prune:true, interval 1h
(a CNI does not need a 10m re-render cadence; kustomize-controller is source-triggered so
a push still reconciles immediately — the interval is only the drift-correction floor),
timeout 10m (a 3-node DaemonSet rollout with wait exceeds the 5m default),
dependsOn:[flux-system], and postBuild.substituteFrom: cluster-vars — the last is
LOAD-BEARING, not decorative: helmrelease.yaml references ${K3S_API_HOST} and
kustomize-controller now runs --feature-gates=StrictPostBuildSubstitutions=true, so
without it this Kustomization would fail immediately.
*** THE HAZARD WAS FAR WORSE THAN "the CNI gets pruned" ***
platform/cilium/namespace.yaml is the ONLY declaration of Namespace/kube-system anywhere
under ./platform, and the live object carries kustomize.toolkit.fluxcd.io/name=platform.
With prune:true, the moment ./platform/cilium leaves platform's path the stale set is not
"the Cilium HelmRelease" — it is `kubectl delete namespace kube-system`, taking coredns,
metrics-server, local-path AND Cilium. A full cluster kill, not a CNI kill.
TWO INDEPENDENT PROTECTIONS, deliberately not trusting either alone:
1. prune:false on `platform` for this transition merge — Flux's documented Procedure 1 for
moving resources between Kustomizations. Chosen over the per-resource
`kustomize.toolkit.fluxcd.io/prune: disabled` annotation because that needs THREE merges
and leaves the new Kustomization owning resources it can never prune.
2. The pruner's own owner-label filter. In v1.4.0, prune() passes Inclusions:
GetOwnerLabels(obj.Name...), so platform only deletes objects whose LIVE labels still
say name=platform. The new cilium Kustomization relabels all three to name=cilium on
apply, and platform.dependsOn:[cilium] is REVISION-GATED, so platform provably cannot
reconcile revision N until cilium has APPLIED revision N. The relabel deterministically
precedes the prune.
Protection 2 alone would probably have sufficed. Not betting kube-system on "probably".
NOTED FOR THE FUTURE, in-file: anyone moving a platform/* component must first check
whether it is the sole declarer of a shared Namespace — that turns a component-scoped
prune into a cluster-scoped one.
Verified: `kubectl kustomize platform` goes 88 -> 85 documents, 3 removed / 0 added /
0 CHANGED, and `kubectl kustomize platform/cilium` reproduces those 3 byte-identical
(sha256-matched). A pure ownership move; no Cilium value touched. L0 substitution gate
103 assertions (was 102), new cilium row, 0 hazards. Server-side dry-run against the live
cluster accepted both CRs. clusters/prod renders clean off the same base.
MERGE 2 (restore prune:true) follows only after verifying: 13/13 Ready; kube-system, the
HelmRelease and the HelmRepository all relabelled to name=cilium; platform's inventory no
longer lists them; and Cilium pod ages/restarts UNCHANGED from
cilium-2f8h8 r=0 2026-09-01T05:55:27Z / cilium-425m8 r=1 2026-08-30T13:25:10Z /
cilium-54cww r=0 2026-09-01T05:55:26Z, HelmRelease v11.
Serves: R-PLAT-01, R-GITOPS-01.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
During the substitution incident a bad render (k8sServiceHost: "") took down TWO of three
Cilium agents simultaneously. The agents' `config` init container falls back to the API
ClusterIP, which only the Cilium agent itself can translate — and that agent is the one
being replaced — so it deadlocks on `dial tcp 10.43.0.1:443: i/o timeout`.
Root cause of the 2-at-once: `updateStrategy.rollingUpdate.maxUnavailable: 2` is the
UPSTREAM CHART DEFAULT, confirmed by downloading cilium-1.20.1.tgz and reading
values.yaml:437-444 — not a local choice, and not overridden anywhere in this repo.
maxUnavailable: 1 makes a rollout replace ONE agent at a time and stop at the first pod
that fails to become Ready. The whole value of the setting is that the rollout halts on
the first bad node instead of taking the majority of the dataplane with it, and that
guarantee only exists at 1. It matters far more on a cold PROD bootstrap, where there is
no healthy peer to recover from — dev only self-healed because k3s-dev-3 kept a working
agent throughout.
Set as a [parity] field in ALL THREE places, per this repo's documented contract that the
Ansible bootstrap values and the Flux HelmRelease values must agree on settings that
define cluster networking (Ansible installs Cilium at bootstrap; the HelmRelease adopts
it; the two layers fight otherwise):
platform/cilium/helmrelease.yaml (authoritative for Flux)
infra/ansible/roles/cilium/templates/cilium-values.yaml.j2 (authoritative for bootstrap)
platform/cilium/values.yaml (reference mirror, kept in sync)
Verified by rendering the .j2 with roles/cilium/defaults and diffing full value sets: every
[parity] field is byte-equal across all three; only k8sServiceHost differs, as designed
(${K3S_API_HOST} vs {{ k3s_api_host }}).
MERGING THIS IS NON-DISRUPTIVE. updateStrategy lives at DaemonSet.spec.updateStrategy — a
SIBLING of spec.template (chart daemonset.yaml:49-53). The DaemonSet controller decides a
pod is out of date by hashing spec.template alone, so the pod-template hash does not move,
no new ControllerRevision is created, and zero pods are recreated. Helm patches the field
in place. (If it were disruptive it would be self-defeating: rolling agents under the old
maxUnavailable: 2 in order to install maxUnavailable: 1.) Verify after reconcile with
`kubectl -n kube-system get ds cilium -o jsonpath='{.spec.updateStrategy.rollingUpdate.maxUnavailable}'`
plus unchanged pod ages — that is the cleanest proof nothing rolled.
Trade-off recorded in the comments: upgrades serialise, roughly two waves to three at 3
nodes (a minute or two). On a large cluster this becomes painful and a percentage such as
"10%" would be right — noted so a future larger cluster revisits it deliberately.
SCOPE: this bounds blast radius, it does NOT close the substitution race. A bad render
still ships; it now breaks one node and stalls rather than taking the majority.
platform.dependsOn: [flux-system] (P0) is the fix for the cause.
Serves: R-PLAT-01, R-NET-01.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Adds conformance/tests/l0-static/flux-postbuild-substitution/ — an offline, stdlib-only
L0 check tagged R-ENV-02 + R-GITOPS-01 (deliberately NOT R-TEST-01, which would
overclaim). It reuses lib/status.py's `resources:` walk rather than forking it, so "what a
path renders" has exactly one definition. It asserts: every ${...} on a substituted path
is defined-or-escaped (comments included); no default-operator form on an undefined name;
no unparseable expansion; the rendering path actually substitutes cluster-vars; every
overlay-patched /data/KEY has a base default; and that the gate itself is load-bearing.
WIRED AS AN UNCONDITIONAL STAGE (3c) in ci/l0.sh, and that matters: stage 3b scopes to
touched requirements, and ci/touched-requirements.sh maps only clusters/* to R-ENV-02 —
so an edit to a substituted manifest under apps/ or platform/ would have filtered this
check out of exactly the change that introduces the bug. l0 fail -> BLOCK=1 in gate.sh.
Mutation-tested all five assertion classes (undefined token, indexed expansion, :- form,
missing substituteFrom, un-wired gate); each fails as intended, baseline green.
STRICT MODE IS PROVEN SAFE TO ENABLE — measured, not inspected. For all 11 substituted
paths, `kubectl kustomize --load-restrictor LoadRestrictionsNone <path> | flux envsubst
--strict` exits 0 against cluster-vars base defaults. Negative control confirms the
incident: with K3S_API_HOST unset, non-strict renders k8sServiceHost empty while strict
exits 1 with `variable not set (strict mode)`. The feature gate itself is NOT flipped
here — that is a live-cluster change and the operator's call.
Escapes the six comment hazards that stood in the way (apps/vms/unifi x4,
platform/longhorn x2). All were prose ABOUT the escaping convention; since the repo's own
contract is that $${...} emits a literal ${...}, escaping them is both precise and
self-consistent. kustomize strips these particular comments today, but comment survival is
an implementation detail — a body comment carrying $ WAS observed surviving into built
apps/wave-1 — and a survivor would kill the whole Kustomization.
DOCTRINE CORRECTED (measured against flux 2.4.0): clusters/base/apps-vms.yaml claimed an
indexed expansion is fatal "escaped or not". Escaping is TOTAL — $${ARR[0]} emits the
literal text safely; only the UNESCAPED indexed form is fatal. Corrected in place.
Also empties DOCUMENTED_GAPS in conformance/run.sh with rationale: R-ENV-02 and
R-GITOPS-01 now have a real verifying test, which run.sh's own comment named as the
trigger for removing them.
Advisory output (not failures): 6 dead cluster-vars knobs — BASE_DOMAIN, CLUSTER_ENV,
CLUSTER_POD_CIDR, PIA_REGION, PLATFORM_NODE_SELECTOR_{KEY,VALUE}. The last four are
documented reference-only; BASE_DOMAIN and CLUSTER_ENV look like genuine rot (hostnames
are hardcoded elsewhere) and are worth a follow-up.
Gate: MERGE GATE: PASS. 102 assertions, 0 hazards.
Serves: R-ENV-02, R-GITOPS-01.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
R-ENV-02 promises dev and prod differ ONLY by scoped cluster-vars overrides. In fact
dev LAN addresses were hardcoded into base manifests, so clusters/prod would have come
up ANNOUNCING DEV ADDRESSES (192.168.1.x) on the prod LAN (192.168.2.x).
New cluster-vars knobs (base default = dev literal, per the file's existing convention):
K3S_API_HOST platform/cilium/helmrelease.yaml (+ the values.yaml mirror)
CILIUM_LB_POOL_START platform-config/cilium/l2-ip-pool.yaml
CILIUM_LB_POOL_STOP "
UNIFI_L4_VIP apps/vms/unifi/50-service-l4.yaml
CRAFTY_L4_VIP platform/crafty/service-l4.yaml
S3_BACKUP_ENDPOINT platform/keycloak/postgres-cluster.yaml, apps/wave-0/{nextcloud,
semaphore}-db.yaml, and platform/velero/helmrelease.yaml
Also fixes GATEWAY_VIP's base default, which was still the RETIRED libvirt address
10.73.0.240, and adds the GATEWAY_VIP patch that clusters/prod was missing entirely.
clusters/base/apps-vms.yaml gains postBuild.substituteFrom (it was the one consuming
Kustomization without it, so a ${TOKEN} there would have shipped as a literal string).
The envsubst contract is documented there: ${VAR} substitutes and escapes as $${VAR};
$VAR and $(cmd) are left alone; ${ARR[0]} hard-errors. The three UOS_URL expansions in
the unifi cloud-init are escaped accordingly.
PROD ADDRESS PLAN (operator, 2026-08-31):
nodes .10/.11/.12 · control-plane VIP .9 (Ansible-side) · GATEWAY_VIP .20
Cilium LB pool .20-.30
CONTAINMENT INVARIANT, now documented at the pool and in the prod overlay: GATEWAY_VIP
MUST equal CILIUM_LB_POOL_START. The shared Gateway pins no address, so Cilium LB-IPAM
gives it the first free address of a matching pool block. An initial prod proposal put
.20 outside a .21-.30 pool; the Gateway would have come up on .21 while split-horizon
coredns-custom kept answering *.home.keithjackson.dev with .20 — an address nothing
announces — blackholing in-cluster OIDC hairpin traffic while reporting healthy. The
pool was widened to .20-.30. The control-plane VIP (.9) stays outside the pool so
kube-vip and LB-IPAM can never contend.
Values prod still cannot know stay INVALID `TBD-PROD-*` placeholders on purpose: base
defaults are dev values, so an omitted patch means prod silently inherits a dev address
— far worse than a loud failed apply.
DEV IS BYTE-IDENTICAL. Verified by kustomize build over 13 paths before/after, then
flux envsubst with the dev cluster-vars values applied to both sides: zero drift. The
only deltas are the intended additive keys in the overlay ConfigMap itself. No
unresolved ${TOKEN} survives. render-golden green (10 goldens); yamllint clean.
Serves: R-ENV-02. Task: T-7.1 (prerequisite). Gate: MERGE GATE: PASS.
Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
The kustomization pulled two REMOTE github bases (?ref=v8.2.0) that Flux's kustomize-controller
git-fetched on every reconcile; those fetches intermittently exceeded the 27s timeout ->
'kustomize build failed' BuildFailed events -> the whole platform tier stalled whenever GitHub was
slow. Vendored the exact v8.2.0 output to local crds.yaml + snapshot-controller.yaml (built with
kubectl kustomize; images transformer still re-pins the controller to v8.2.0), so the build is now
offline + deterministic. No change to what's deployed. platform/kubevirt has the same latent pattern.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Roll the verified Option-B per-app oauth2-proxy sidecar out to the remaining
four forward-auth apps after the pihole canary was confirmed fail-closed
(unauthenticated -> 302 to Keycloak, pod 2/2, clean sidecar OIDC discovery).
Each app's Gateway route now terminates at its oauth2-proxy sidecar (:4180,
--allowed-group=media-users, --upstream=http://127.0.0.1:<uiport>); the
gateway-ingress CiliumNetworkPolicy allows the Gateway to reach ONLY :4180, so
the bare UI is never served un-authenticated (fail-closed; R-IDP-02/06). The
apps' own app/API ports stay in-Service for EAST-WEST callers (arr ->
qbittorrent:8080 / jackett:9117) via the existing reciprocal-ingress allows.
- keycloak: confidential jackett/qbittorrent/radarr/sonarr clients (realm-import
+ realm-clients mirror), provisioned live via kcadm (groups membership mapper).
- secrets: four SOPS/age <app>-oauth2-proxy-credentials, wired into apps/wave-0
and apps/wave-1 by the renderer's sidecar-creds wave rule.
- ci: remove the temporary canary deferral (all five goldens now reproduce).
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Close the un-gated-UI security finding for forward-auth apps (R-IDP-02) by
rendering the Option-B per-app oauth2-proxy reverse-proxy sidecar that crafty
proves working. catalog/render.py now emits, for any forward-auth app with an
exposed UI: the oauth2-proxy sidecar (upstream http://127.0.0.1:<uiport>,
:4180, --allowed-group per auth.group), an authproxy :4180 Service port, an
HTTPRoute backendRef -> :4180, and a gateway-ingress CiliumNetworkPolicy that
allows the Gateway to reach ONLY :4180 (fail-closed: the bare UI is never
served un-authenticated). oidc/native/none apps render byte-identical.
This is the CANARY: only pihole's golden carries the sidecar and is verified
fail-closed in-cluster first. jackett/qbittorrent/radarr/sonarr are deferred to
the rollout commit (temporarily listed in ci/render-golden.sh) so their UIs
keep serving exactly as today until their clients+secrets land.
- keycloak: confidential pihole client (realm-import + realm-clients mirror),
provisioned live via kcadm (groups membership mapper, bare names).
- secret: SOPS/age pihole-oauth2-proxy-credentials (client-id/secret/cookie),
wired into apps/wave-0 via the renderer's new sidecar-creds wave rule.
- L1 render goldens regenerated for the sidecar (forward-auth fixtures).
Co-Authored-By: Claude Opus 4.8 <[email protected]>
App SSO broke cluster-wide after the Cilium 1.16.5->1.20.1 upgrade: the CVE-2025-30162 fix (in
>=1.16.8/1.17.2) makes the shared Gateway's L7 proxy re-evaluate a pod's egress policy against
Keycloak's REAL backend cluster identity (enforce_policy_on_l7lb) when it hairpins to keycloak.home.
'toEntities: world' matches the Gateway VIP at L3/L4 but not the resolved keycloak identity at L7, so
OIDC discovery/token 403'd ('Access denied') for every app with a NetworkPolicy (gitea/nextcloud/
semaphore/headlamp/crafty...); Grafana survived only because monitoring has no NetworkPolicy.
Fix: explicit toEndpoints egress allow to the keycloak backend for SSO apps.
- catalog/render.py: emit it for auth.mode in {oidc,forward-auth}; re-rendered 8 app goldens.
- platform/headlamp + platform/crafty: hand-written policies get the same allow.
Split-horizon design unchanged. Verified live: headlamp-ns -> keycloak discovery 403 -> 200.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
The browser-mfa conditional-user-role authenticator used config key 'condition.evaluated.user.role',
which is not a real Keycloak property — the role resolved to null, so Keycloak logged 'Invalid role
name submitted: null' and the REQUIRED conditional-OTP step threw, 403-ing every platform-admins login
(Headlamp, Grafana admin, etc.). Correct key for Keycloak 26 is condUserRole=platform-admin.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Adds NodeCPUPressureHigh (CPU PSI 'some' > 15% for 15m) and NodeCPULoadHigh (load5/core > 1 for 15m)
to platform-health.nodes, plus a 'Nodes — CPU / Memory Pressure' Grafana dashboard (PSI + load-per-core
+ utilization). These surface the bursty CPU contention that averaged CPU%/kubectl top hides — the
signal behind 'spikes cause blips of delay'. Makes the 6->8 vCPU resize measurable.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Both were pinned at ~98% of their limits (Prometheus 2002/2048Mi, Grafana 499/512Mi) with no
headroom for TSDB compaction / dashboard-render spikes — the classic OOM setups. Nodes are 24Gi at
38-44% used (~14Gi free each; the '~22 GiB ceiling' comment predates the 16->24Gi node bump), so the
extra headroom is comfortable. Grafana request 256->384Mi, Prometheus request unchanged.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
chart 7 major bump. Verified chart 7.3.0 renders identically for this SingleBinary/filesystem/
Longhorn config (one loki StatefulSet, no extra read/write/backend/cache pods, no schema errors).
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Text-only cleanup after the upgrade campaign. Updates stale version
strings, prose, and comments to the verified current facts; no
Kubernetes behavior or version pins changed except the kubectl
toolchain bump (below).
Verified facts applied: k3s v1.36.4+k3s1, Debian 13 trixie,
Cilium 1.20.1 / Gateway API v1.6.1, Longhorn 1.12.1, node RAM 24 GiB,
dedicated virtio2 SSD etcd disk per node, Velero chart 12.x / v1.18,
Grafana 13.2.0.
- Version strings across infra/ and platform/ READMEs + comments
- kubectl toolchain 1.31.4 -> 1.36.4 (devcontainer/ci/conformance
versions.env; 1.31 vs a 1.36 cluster exceeded the +-1 skew)
- Retired-single-apps-Kustomization refs -> apps-wave-0/1/2 + apps-vms
with per-wave SOPS decryption (flux/clusters docs + bootstrap.sh)
- COMPLETED/historical banners on the three executed plan docs
- Resolved TODO(WS-SEC) markers -> Renovate-maintained ranges
- Appended dated DECISIONS-LOG entry (2026-08-30) superseding the
2026-08-26 LIVE STATE line
Left as revalidation items (not blind-edited): external-snapshotter
v8.2.0 vs 1.36, Cilium ext_authz/forward-auth relevance under 1.20.1,
conformance ImageVolume feature-gate assertions under 1.36.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Grafana 12.x bundled in kps 88 OOMKilled (exit 137) at 256Mi during grafana-apiserver/unified-storage
startup — no fatal log, just SIGKILL. The new Grafana pod never became Ready, so Helm --wait ran the full
20m and rolled back to 65.8.1 every time. 512Mi holds.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
The 65->88 multi-major upgrade exceeded the default 5m Helm timeout (full operator/KSM/
node-exporter + Prometheus/Alertmanager StatefulSet rollout), so Flux rolled back to 65.8.1
four times and Stalled, blocking the platform kustomization and everything downstream.
CRDs pre-applied server-side at operator v0.93.1.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Root cause of a W5 wedge: default 5m Helm timeout on slow-HDD Longhorn upgrades caused
a rollback attempt after the data had already migrated to the new minor, which the old
manager fatally refuses. Give the upgrade room and never auto-rollback (fail forward).
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Gateway API CRDs bumped v1.2.1->v1.3.0 (required by Cilium 1.18; applied server-side).
Land on current k8s 1.32 before the 1.33 hop.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Land the two pacing components on current k8s 1.31 before the 1.32 k3s hop.
Cilium 1.17 supports k8s 1.29-1.32; Longhorn 1.8 supports 1.25+; Gateway API CRDs
stay v1.2.1 (Cilium 1.17 targets v1.2.x). Also commits the full upgrade plan doc.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Migrate image.toolkit.fluxcd.io/v1beta2 -> /v1 for ImageRepository,
ImagePolicy, and ImageUpdateAutomation ahead of the Flux v2.7+ upgrade
(cluster-upgrade-plan §3.2). Flux v2.4 (current) already serves /v1, so
migrating the manifests now is a safe no-op at runtime.
- catalog/render.py: r_image_automation emits v1 for ImageRepository/ImagePolicy
- platform/image-automation/{image-update-automation,per-service/pihole.example}
- conformance L3 hands-off-update chainsaw applies v1 CRs
- conformance L1 goldens re-rendered to match (apiVersion-only diff)
No spec-field changes required: all spec fields used (image, interval,
imageRepositoryRef, filterTags.pattern/extract, policy.semver.range,
sourceRef, git.*, update.strategy/path) are unchanged in v1; the v1
messageTemplate already uses .Changed (not the removed .Updated).
apps/ golden unaffected (canonical render uses --no-image-automation).
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Headlamp OIDC login failed with `unsupported protocol scheme ""`. The
kubernetes-sigs/headlamp chart `envFrom`s the headlamp-oidc Secret verbatim
and renders args like `-oidc-idp-issuer-url=$(OIDC_ISSUER_URL)`, so the Secret
keys must be the exact env var names. The keys were camelCase
(clientID/issuerURL/scopes/callbackURL/clientSecret), following the mistaken
`.example`, so `OIDC_ISSUER_URL` was never defined. Kubernetes left the literal
`$(OIDC_ISSUER_URL)` in the arg and Headlamp's OIDC discovery hit
`Get "$(OIDC_ISSUER_URL)/.well-known/openid-configuration": unsupported
protocol scheme ""`.
Rename the SOPS Secret keys to OIDC_CLIENT_ID / OIDC_CLIENT_SECRET /
OIDC_ISSUER_URL / OIDC_SCOPES / OIDC_CALLBACK_URL (values unchanged, still
SOPS/age-encrypted). Fix the `.example` and README so the correct, load-bearing
key names are documented and the camelCase footgun can't recur.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
The k3s VM disks are HDD-backed, so a full Longhorn replica rebuild's
sequential I/O can saturate the spindle and starve etcd's WAL fsync,
crash-looping etcd and flapping the kube-vip VIP (observed during a
rolling node resize). Cap rebuilds to one replica per node at a time so a
node bounce or replica replenishment can never storm the disks.
Set via HelmRelease defaultSettings.concurrentReplicaRebuildPerNodeLimit
(GitOps-managed, permanent). replica-replenishment-wait-interval stays at
the Longhorn default of 600s. Reference values.yaml kept in sync.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Supersedes the previous kubectl-image repoint. The velero chart's pre-install
`velero-upgrade-crds` HOOK job runs docker.io/bitnami/kubectl, whose public
Docker Hub catalog Bitnami removed (Aug 2025); `bitnami/kubectl:1.31` 404s,
leaving the hook in Init:ImagePullBackOff so Velero never installs.
Repointing the hook to the exact-version rancher/kubectl:v1.31.5 (prior commit)
does not work: that image is distroless, and the hook's init runs
`/bin/sh -c "cp $(which sh)/... $(which kubectl) /tmp"` — with no /bin/sh it
crashes StartError (verified on-cluster).
The hook exists ONLY to UPDATE CRDs on chart upgrades — it is NOT how CRDs are
first installed. The chart ships all 13 Velero CRDs in its native `crds/`
directory, which Helm applies unconditionally on install (before hooks),
independent of this flag; the CRDs are already present on-cluster from Helm's
crds/ handling. Setting `upgradeCRDs: false` drops the fragile external
kubectl-image dependency entirely while CRDs still install. Trade-off: Helm
won't manage crds/ after install, so CRD schema upgrades become a
manual/Renovate step; acceptable given Velero's CRDs are stable.
All other Velero values (BSL -> versitygw s3://velero, kopia FSB node-agent, the
two Schedules, the velero-cloud-credentials secret) are unchanged.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
The velero chart's pre-install `velero-upgrade-crds` hook defaults its kubectl
image to docker.io/bitnami/kubectl, whose public Docker Hub catalog Bitnami
removed (Aug 2025). `docker.io/bitnami/kubectl:1.31` now 404s, leaving the hook
job in Init:ImagePullBackOff so Velero never installs (no Deployment,
node-agent, or BackupStorageLocation).
Set kubectl.image to the Rancher-maintained mirror docker.io/rancher/kubectl,
pinned to the cluster's exact k8s minor v1.31.5 (multi-arch amd64/arm64,
verified pullable via the Docker Hub registry API). All other Velero values
(BSL → versitygw s3://velero, kopia FSB node-agent, the two Schedules, the
velero-cloud-credentials secret) are unchanged.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Stand up a single-node/single-drive MinIO S3 endpoint on the Proxmox host as the
off-cluster backup target for Velero (R-STATE-03) and CloudNativePG (R-IDP-05).
Ansible (author + executed):
- roles/minio_backup_target + playbooks/backup-minio.yml — idempotent, additive only:
ZFS dataset Storage/k3s-backups (zstd, quota=100G, /mnt/Storage/k3s-backups, no
sharenfs); minio system user; pinned+checksummed MinIO/mc binaries (fetched on the
controller since the host has no WAN, copied to host); hardened systemd unit bound to
the host's cluster-reachable LAN IP (auto-discovered) on :9000; velero + cnpg buckets;
a non-root access key scoped by IAM policy to only those two buckets.
- inventory: new [proxmox_hosts] group ([email protected].2.102); overrides the cluster-wide
ansible_user. Does not disturb existing k3s node entries.
- group_vars/proxmox_hosts/vault.yml (ansible-vault): root + scoped MinIO creds.
GitOps hand-off:
- platform/backup/minio-credentials.sops.yaml (+ .example) — SOPS/age-encrypted endpoint +
scoped key/secret for the Velero/CNPG wiring (WS-STORAGE/WS-SEC). Not yet wired into a
kustomization (left for the consuming agents).
Verified live: dataset props correct; minio active+enabled bound to 192.168.1.9:9000;
S3 endpoint http://192.168.1.9:9000/minio/health/ready returns 200 from an in-cluster pod;
buckets velero+cnpg present; scoped key created+policy attached; re-run is idempotent.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
The plural --allowed-groups does not exist in oauth2-proxy v7.8.1; the shared
oauth2-proxy new pod and the crafty auth sidecar both crashlooped on 'unknown flag'.
The flag is the singular, repeatable --allowed-group.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Add Crafty Controller 4 (registry.gitlab.com/crafty-controller/crafty-4:4.10.8)
as a platform component: a web panel that runs Java + Bedrock Minecraft servers as
subprocesses in one pod + a 20Gi Longhorn PVC (mounted 5x by subPath).
Auth = Option B (per-app oauth2-proxy reverse-proxy sidecar), since Cilium 1.16.5
cannot do ext_authz on the shared Gateway:
- oauth2-proxy v7.8.1 sidecar (:4180) reverse-proxies to Crafty's self-signed HTTPS
UI (127.0.0.1:8443), gated to Keycloak homelab realm group platform-admins.
- Confidential `crafty` client provisioned live via kcadm + captured in the realm
import; client secret + cookie secret via SOPS/age (mirrored in realm-clients).
- Web UI: HTTPRoute crafty.home.keithjackson.dev -> Service :80 -> sidecar :4180.
- Game ports: Cilium L2 LoadBalancer on VIP 192.168.1.55 (next free after .53/.54),
starter set TCP 25565-25567 (Java) + UDP 19132-19134 (Bedrock), NOT auth-gated.
- fsGroup:0 / runAsGroup:0 / runAsUser:1000 for the image's crafty:root identity on
the root:root Longhorn PVC (the seerr fsGroup lesson).
- Own non-gating Flux tier clusters/base/crafty.yaml (dependsOn platform-config,
wait:false) + sops-decryption patch in clusters/dev — a leaf UI must never gate
the platform.
- Explicit default-deny NetworkPolicy + CiliumNetworkPolicies (platform namespace,
not app-namespace, so Kyverno does not auto-generate the baseline).
make -C ci gate green; kustomize build of platform/crafty, clusters/base,
clusters/dev clean.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Build the deferred forward-auth ENFORCEMENT for non-OIDC UIs (pihole/*arr/
qbittorrent/homepage; Crafty next), gated by Keycloak via oauth2-proxy.
- helmrelease: add `--allowed-groups=platform-admins` (R-IDP-06 floor:
authenticated AND in group) + `--skip-provider-button` (302 straight to
Keycloak; verified the interstitial 403 sign-in page otherwise).
- forward-auth-cec.yaml: interim self-authored CiliumClusterwideEnvoyConfig
ext_authz listener -> oauth2-proxy /oauth2/auth, failure_mode_allow:false,
per-route opt-in via virtualHost. TEMPLATE — not Flux-wired (PM deploys with
an L2 VIP + DNS repoint). The ext_authz stanza is verified in-cluster.
- README: document the mechanism, why per-route ext_authz on the shared Gateway
is NOT clean on Cilium 1.16.5 (machine-owned cilium-gateway-shared CEC; native
ExternalAuth/GEP-1494 in no released Cilium), the three options + recommendation,
the opt-in steps, the in-cluster proof, and the human browser test.
Keycloak `oauth2-proxy` client already carries the callback redirect URI + a
bare `groups` claim (full.path=false) — no realm change needed.
Proof (cilium-envoy v1.30.8): a standalone envoy running this exact ext_authz
stanza against the live oauth2-proxy denied unauthenticated requests (403,
oauth2-proxy sign-in body; upstream never served), while a no-filter control
route on the same backend returned 200. R-IDP-02 property holds.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
The 7 falsely-firing platform alerts (KeycloakDown, KeycloakDatabaseDown,
KubeAPIDown, KubeletDown, VPNTunnelDown, KubeVirtControlPlaneDown,
CDINoDefaultStorageClass) share ONE root cause, not per-rule label mismatches:
the Prometheus PVC is 100% full ("no space left on device" on every WAL commit).
Evidence from the live dev cluster:
- /prometheus is 4.8G/5Gi, 0 bytes free; every scrape logs "Scrape commit
failed ... write to WAL: no space left on device". Scrapes still return 200
(targets show healthy, lastScrape=now) but samples are never stored, so the
TSDB head froze at 2026-08-28 13:45 UTC while the pod clock reads ~19:33 UTC.
- Instant queries at "now" therefore return an EMPTY vector, so every
absent()/==0 platform rule trips. Evaluated at the last-good-sample time each
expr is correctly FALSE (keycloak-0/keycloak-db-1 ready=1, apiserver+kubelet
up=1 x3, tun0 series present, kubevirt control-plane expr empty, CDI default
present) — the rules are correct; the monitor was simply blind.
`retention` alone prunes only by AGE: this cluster is 3d2h old with 3d dev
retention, so time-pruning has NEVER run and 3d of block churn (~1.2Gi/day) plus
a WAL that ballooned past 1Gi overran the 5Gi disk. Fix adds a size-based
retention cap (new MONITORING_PROMETHEUS_RETENTION_SIZE knob) kept below each
env's PVC so Prometheus drops oldest blocks before the disk fills, and bumps the
dev PVC 5Gi->8Gi for real WAL/head/compaction headroom.
- platform/monitoring/kube-prometheus-stack.yaml: retentionSize: ${knob}
- clusters/base/cluster-vars.yaml: knob default 8GiB (base 10Gi PVC)
- clusters/dev/kustomization.yaml: PVC 5Gi->8Gi, retentionSize 5GiB
- clusters/prod/kustomization.yaml: retentionSize 45GiB (50Gi PVC)
- platform/monitoring/README.md: knob table + rationale
Deploy note (PM): the PVC is already 100% full, so Prometheus cannot self-prune
until it can write. The 5Gi->8Gi bump must land as a Longhorn online expansion
(operator PVC expansion / manual patch); once free space returns, retentionSize
keeps it healthy. Watchdog left untouched.
Co-Authored-By: Claude Opus 4.8 <[email protected]>
Plex Deployment had no Plex env vars. Add via the ServiceDefinition env list
(rendered verbatim by catalog/render.py):
- PLEX_UID/PLEX_GID=1000 (image chowns /config to this and drops; matches the
Longhorn ownership convention seerr/semaphore use)
- TZ=Etc/UTC (no other service sets TZ repo-wide)
- ALLOWED_NETWORKS=10.0.0.0/8,192.168.0.0/16 so in-cluster clients (Seerr ->
plex:32400) are trusted as LAN (pod CIDR 10.0.0.0/16 + svc 10.43.0.0/16 + LAN)
PLEX_CLAIM intentionally not set (4-min TTL is unusable via git; one-time live
claim persists on /config). ADVERTISE_IP deferred: the only candidate host
resolves via split-horizon DNS to the internal Gateway VIP and its remote
reachability is unverified — a wrong value breaks remote access, so omitted
per prefer-safety, documented as a follow-up in catalog/plex.yaml.
Homepage still labeled seerr as "Overseerr"; renamed the tile to "Seerr"
(icon mdi-play-box-multiple keeps — not overseerr-specific).
Co-Authored-By: Claude Opus 4.8 <[email protected]>
