Files
home-network-k3s-impl/infra/ansible/roles/k3s_server/tasks/main.yml
T
jakman85andClaude Opus 5 05baae90ec feat(prod): kernel fix + disk layout — prod nodes ready for k3s
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]>
2026-09-04 00:39:10 -05:00

161 lines
7.3 KiB
YAML

---
# Install k3s as an HA server (embedded etcd) with Cilium-required components
# disabled (ADR-0001, docs/02 stage 3 "Cluster-join"). The primary runs
# --cluster-init; secondaries join it with the shared node-token.
- name: Download the k3s install script
ansible.builtin.get_url:
url: "{{ k3s_install_url }}"
dest: "{{ k3s_install_script }}"
mode: "0755"
# --node-ip uses k3s_node_ip (the real 192.168.1.x), NOT ansible_host — the latter is
# now the pfSense NAT SSH target (192.168.2.102:225x) and must not be advertised to k3s.
- name: Compose the common k3s server flags
ansible.builtin.set_fact:
k3s_common_args: >-
server
--flannel-backend=none
--disable-kube-proxy
--disable-network-policy
--disable={{ k3s_disable }}
--write-kubeconfig-mode=644
{% for san in k3s_tls_sans %}--tls-san={{ san }} {% endfor %}
--node-ip={{ k3s_node_ip }}
--data-dir={{ k3s_data_dir }}
# On an ALREADY-installed cluster the install task above is idempotent (creates:
# k3s.yaml) and won't re-apply new --tls-san flags. Persist the SAN list in
# /etc/rancher/k3s/config.yaml, which a running k3s reads on the next restart —
# it then regenerates the serving cert when the SAN set changes, and regenerates
# its containerd config.toml (so nonroot-devices takes effect). Apply the restart
# ROLLING, one node at a time (kube-vip keeps the API reachable; etcd quorum holds
# at 2/3). See playbooks note: run with `serial: 1`.
- name: Ensure /etc/rancher/k3s exists (a freshly re-provisioned node has no k3s dir until install)
ansible.builtin.file:
path: /etc/rancher/k3s
state: directory
owner: root
group: root
mode: "0755"
- name: Write /etc/rancher/k3s/config.yaml (tls-san + nonroot-devices + oidc)
ansible.builtin.copy:
dest: /etc/rancher/k3s/config.yaml
owner: root
group: root
mode: "0600"
content: |
# Managed by Ansible (roles/k3s_server). Read by k3s in addition to the
# systemd unit flags; tls-san entries are merged into the serving cert.
tls-san:
{% for san in k3s_tls_sans %}
- "{{ san }}"
{% endfor %}
# Let non-root pods own the raw block devices bind-mounted into them —
# sets containerd's `device_ownership_from_security_context = true`
# (k3s >= v1.31.3+k3s1 exposes this as the --nonroot-devices flag).
#
# REQUIRED for KubeVirt/CDI on Block-mode PVCs (ADR-0005): CDI's importer,
# uploader and virt-launcher pods run non-root (UID 107, runAsNonRoot,
# cap-drop ALL) and open the volume's raw block device (/dev/cdi-block-volume,
# then the guest disk). Kubernetes only applies fsGroup ownership to
# *filesystem* volumes — it never chowns a raw block device — and Longhorn's
# CSIDriver (fsGroupPolicy: ReadWriteOnceWithFSType) skips fsGroup on RWX/Block
# volumes anyway, so the device stays root:root and the pod fails with
# `blockdev: cannot open /dev/cdi-block-volume: Permission denied`. This flag
# makes containerd set the in-container device node's owner from the pod's
# securityContext, which is the fix the CDI project documents for this case
# (doc/block_cri_ownership_config.md). Node-scoped and security-neutral: the
# pod stays non-root, only the device node it already mounts follows its UID/GID.
# Preserves the longhorn-vm RWX + Block + migratable disk contract unchanged.
nonroot-devices: true
# OIDC (Keycloak SSO for kubectl). The API server trusts ID tokens from the homelab realm;
# the `kubernetes` PUBLIC client (PKCE) + kubelogin mint them, and RBAC binds the `groups`
# claim (platform/keycloak/k8s-oidc-rbac.yaml: oidc:platform-admins->cluster-admin,
# oidc:viewers->view). Nodes resolve keycloak.home.keithjackson.dev to the Cilium gateway VIP
# (192.168.1.53) and trust its public LE-prod cert, so NO oidc-ca-file is needed. The `oidc:`
# prefixes namespace federated identities away from cert / service-account subjects.
# ADDITIVE + break-glass-safe: the cert-based admin kubeconfig keeps full access with or
# without Keycloak, so a bad realm/OIDC state can never lock you out of the cluster.
kube-apiserver-arg:
- "oidc-issuer-url=https://keycloak.home.keithjackson.dev/realms/homelab"
- "oidc-client-id=kubernetes"
- "oidc-username-claim=email"
- "oidc-username-prefix=oidc:"
- "oidc-groups-claim=groups"
- "oidc-groups-prefix=oidc:"
notify: Restart k3s
# --- primary: found the cluster --------------------------------------------
- name: Install k3s primary (cluster-init, embedded etcd)
ansible.builtin.command:
cmd: "{{ k3s_install_script }} {{ k3s_common_args }} --cluster-init"
creates: /etc/rancher/k3s/k3s.yaml
environment:
INSTALL_K3S_VERSION: "{{ k3s_version }}"
when: k3s_server_mode == 'primary'
- name: Wait for the k3s node-token to exist (primary)
ansible.builtin.wait_for:
path: /var/lib/rancher/k3s/server/token
timeout: 180
when: k3s_server_mode == 'primary'
- name: Read the k3s node-token (primary)
ansible.builtin.slurp:
src: /var/lib/rancher/k3s/server/token
register: k3s_token_raw
when: k3s_server_mode == 'primary'
- name: Publish the node-token as a fact for secondaries (primary)
ansible.builtin.set_fact:
k3s_node_token: "{{ k3s_token_raw.content | b64decode | trim }}"
when: k3s_server_mode == 'primary'
# --- secondaries: obtain the join token, then join the existing cluster -----
# Fetch the token straight from the primary via delegate_to, falling back to it
# only when the primary play didn't publish the fact this run. This makes a
# single-node rejoin (`--limit k3s-dev-N`, primary NOT in the batch) self-
# contained — delegate_to still reaches the primary even when it isn't a play
# target, so the join no longer depends on a fact an earlier play had to set.
# T-1.5 finding F2 (see DECISIONS-LOG.md).
#
# "The primary" is resolved through k3s_primary_group (group_vars: `k3s_primary` for
# dev, `k3s_prod_primary` for prod), never a literal group name — so a prod secondary
# delegates to the PROD primary using this same role, unchanged (R-ENV-02).
- name: Fetch the k3s node-token directly from the primary (secondary)
ansible.builtin.slurp:
src: /var/lib/rancher/k3s/server/token
delegate_to: "{{ groups[k3s_primary_group][0] }}"
register: k3s_primary_token_raw
when:
- k3s_server_mode == 'secondary'
- hostvars[groups[k3s_primary_group][0]].k3s_node_token is not defined
- name: Resolve the join token (secondary)
ansible.builtin.set_fact:
k3s_join_token: >-
{{ hostvars[groups[k3s_primary_group][0]].k3s_node_token
| default(k3s_primary_token_raw.content | b64decode | trim, true) }}
when: k3s_server_mode == 'secondary'
- name: Install k3s secondary (join as server)
ansible.builtin.command:
cmd: >-
{{ k3s_install_script }} {{ k3s_common_args }}
--server https://{{ k3s_api_host }}:{{ k3s_api_port }}
creates: /etc/rancher/k3s/k3s.yaml
environment:
INSTALL_K3S_VERSION: "{{ k3s_version }}"
K3S_TOKEN: "{{ k3s_join_token }}"
when: k3s_server_mode == 'secondary'
- name: Wait for the k3s API to be serving locally
ansible.builtin.wait_for:
host: "127.0.0.1"
port: "{{ k3s_api_port }}"
timeout: 180