Files
computer-playbook/cli/deploy/development/inventory.py
T
kevinveenbirkenbachandClaude Opus 4.7 ed04dad5d5 refactor(meta): land req-008/009/010 role-meta layout/networks/ports/run_after migration
Single atomic landing for three back-to-back role-meta refactor requirements
(see docs/requirements/008-role-meta-layout.md, 009-per-role-networks-and-ports.md,
010-role-meta-runafter-lifecycle-migration.md).

req-008 — role meta layout
  * Move every role's metadata under meta/<topic>.yml with file-root
    convention (no wrapping key matching the basename):
      - schema/main.yml + config/main.yml.credentials → meta/schema.yml
      - users/main.yml                                → meta/users.yml
      - config/main.yml.server                        → meta/server.yml
      - config/main.yml.rbac                          → meta/rbac.yml
      - config/main.yml.compose.services + every other top-level key
        of config/main.yml inlined under <entity>.<key>
                                                      → meta/services.yml
      - config/main.yml.compose.volumes               → meta/volumes.yml
  * Drop the legacy schema/, users/, config/ directories from every role.
  * meta/schema.yml: support nested credential keys, optional `default:`
    field (Jinja literal, not rendered/validated/regenerated), implicit
    `algorithm: plain`.
  * Materialised paths drop the `compose.` prefix:
      compose.services.<e>.<…> → services.<e>.<…>
      compose.volumes.<k>      → volumes.<k>
      <former-top-level>.<…>   → services.<entity_name>.<former-top-level>.<…>
  * Update every Python consumer (utils/cache, utils/manager, utils/applications,
    plugins/lookup, plugins/filter, cli/*) plus every Jinja template, vars/*.yml
    and task file to the new paths.
  * No legacy fallback. Repository-wide grep for `compose.services.` /
    `compose.volumes.` / `schema/main.yml` / `users/main.yml` / `config/main.yml`
    is empty outside docs/requirements/ and tasks/utils/migrate_meta_layout.py.
  * Add tests/lint/repository/test_role_meta_layout.py guarding against the
    old paths returning, and an explicit lint guard that meta/variants.yml
    files do not carry a `compose:` wrapper.

req-009 — per-role networks/ports
  * Move group_vars/all/08_networks.yml.defaults_networks.local.<role>.{subnet,
    dns_resolver} into roles/<role>/meta/server.yml.networks.local.{subnet,
    dns_resolver}.
  * Delete group_vars/all/09_ports.yml entirely; ports now live on the
    service entity that exposes them at
    roles/<role>/meta/services.yml.<entity>.ports.{inter,local,public}.
  * Rename + int-cast every former top-level `services.<entity>.port`
    field to `services.<entity>.ports.inter` (single int).
  * Always-category-keyed maps for ports.local / ports.public; relay
    spans use {start, end} integers.
  * Add PORT_BANDS map to group_vars/all/08_networks.yml — single source
    of truth for the per-<scope>.<category> band ranges, consumed by both
    the new CLI helpers and the lint test.
  * New CLI helpers (cli meta ports suggest, cli meta networks suggest)
    walk the live role tree gap-first then increment-fallback, with
    deterministic output and non-zero exit on capacity overflow.
  * cli create role consumes both helpers when scaffolding a new role.
  * Lint extensions: bare-int port shape rejection, host-bound port
    collision check (single ints + relay span integers, inter excluded),
    band-membership check with documented allow-list for the legacy
    web-app-bigbluebutton.bigbluebutton.local.http: 48087 entry.

req-010 — run_after / lifecycle migration
  * Move galaxy_info.run_after and galaxy_info.lifecycle out of
    meta/main.yml into meta/services.yml.<primary_entity>.{run_after,
    lifecycle}; <primary_entity> is the value returned by
    get_entity_name(role_name).
  * For multi-entity roles whose primary entity has no compose service
    (web-app-bluesky → bluesky, web-app-matrix → matrix, desk/dev/drv/
    gen roles), create a dedicated metadata-only top-level entry under
    <primary_entity>.
  * Introduce utils/roles/meta_lookup.py with get_role_run_after /
    get_role_lifecycle that resolve <primary_entity> once and read from
    meta/services.yml; degrade gracefully when meta/services.yml is
    absent OR the field is absent; raise MetaServicesShapeError when the
    file exists but is malformed.
  * Update every consumer (utils/service_registry, utils/invokable,
    cli/meta/applications/resolution/run_after/__main__, cli/meta/roles/
    lifecycle_filter/__main__, cli/meta/applications/resolution/combined,
    cli/build/{tree,graph,role_include}, cli/deploy/development/{common,
    init,deps,deploy}, plugins/filter/canonical_domains_map, etc.) to
    use the helper.
  * Lint guards: no run_after/lifecycle in meta/main.yml; lifecycle
    values restricted to {planned,pre-alpha,alpha,beta,stable,deprecated};
    run_after/lifecycle never on a non-primary entity; empty run_after
    lists rejected.

Also folded into this change set
  * services_disabler now writes `services.<svc>.{enabled,shared}` into
    host_vars instead of the old `compose.services.<svc>.{...}` path; the
    discover/find pair walks meta/services.yml and only treats entries
    carrying enabled/shared/provides as provider declarations (so
    metadata-only entries from req-010 do not create spurious provider
    rows).
  * service_registry skips `run_after` deps whose target was filtered
    out via SERVICES_DISABLED instead of failing the whole load.
  * cli/mirror/resolver and cli/create/inventory/mirror_overrides emit/
    consume the new applications.<role>.services.<svc>.image shape
    (mirror.md updated to match).
  * utils/domains/list.build_applications_from_roles now reads
    meta/server.yml directly (it had been silently returning {} since
    config/main.yml stopped existing).
  * utils/roles/resource_validation.filter_roles_by_min_storage reads
    meta/services.yml.<entity>.min_storage instead of the legacy
    compose.services path.
  * web-app-bluesky pre-clones the social-app build context BEFORE
    sys-svc-compose so the build context is on disk by the time
    sys-stk-front-proxy/tasks/main.yml flushes the docker build handler.
  * 4 meta/variants.yml files (svc-ai-ollama, web-app-phpmyadmin,
    web-app-wordpress, web-app-keycloak) lose their `compose: services:`
    wrapper so deep-merge over meta/services.yml lands on the new
    file-root shape.

Validation
  * Unit tests:        1571 passed.
  * Lint tests:        50 passed.
  * Integration tests: 161 passed (1 skipped).
  * Full-cycle fresh-purged-apps deploy succeeded for the 8-app validation
    set: web-app-{nextcloud,wordpress,yourls,bigbluebutton,bluesky,gitea,
    discourse,mattermost}.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-04-27 09:58:26 +02:00

291 lines
10 KiB
Python

"""SPOT for development inventory creation.
Anything that wants to materialise the development inventory MUST go
through `build_dev_inventory` (single folder) or
`build_dev_inventory_matrix` (one folder per matrix-deploy round). Both
flow through `DevInventorySpec` and own:
* the SPOT vars-file (`DEV_INVENTORY_VARS_FILE`),
* per-app variant baking (each role's `meta/variants.yml` is resolved at
build time and emitted as `applications.<app>` overrides into the
inventory's `host_vars`, with variant 0 as the fallback for apps
without `meta/variants.yml`),
* the vault password file.
The deploy wrapper consumes `plan_dev_inventory_matrix` to know which
folder to deploy against in each round; the planner is a pure function
so wrapper and init step compute the same plan independently.
"""
from __future__ import annotations
import json
from dataclasses import dataclass
from typing import TYPE_CHECKING, Any, Mapping, Sequence
from utils.cache.applications import get_variants
from .common import DEV_INVENTORY_VARS_FILE
from .mirrors import generate_ci_mirrors_file, should_use_mirrors_on_ci
if TYPE_CHECKING:
from .compose import Compose
@dataclass(frozen=True)
class DevInventorySpec:
"""Everything `infinito create inventory` needs to materialise one
development inventory folder.
`extra_vars` wins over the implicit `STORAGE_CONSTRAINED` / `RUNTIME`
overrides so callers (and tests) keep the documented "user-provided
vars always win" behaviour without re-implementing the merge.
`active_variants` is the per-app variant index for the round this
folder represents. Apps absent from the mapping (or with an out-of-
range index) fall back to variant 0. The resolved variant payload
for every app in `include` is baked into the inventory as an
`applications.<app>` override, so the deploy stage no longer needs
a runtime variant selector — the inventory itself is variant-resolved.
"""
inventory_dir: str
include: tuple[str, ...]
storage_constrained: bool
runtime: str
extra_vars: Mapping[str, Any] | None = None
services_disabled: str = ""
active_variants: Mapping[str, int] | None = None
def __post_init__(self) -> None:
if not self.include:
raise ValueError("DevInventorySpec.include must not be empty")
if not isinstance(self.include, tuple):
object.__setattr__(self, "include", tuple(self.include))
def overrides(self) -> dict[str, Any]:
merged: dict[str, Any] = {
"STORAGE_CONSTRAINED": bool(self.storage_constrained),
"RUNTIME": self.runtime,
}
if self.extra_vars:
merged.update(self.extra_vars)
return merged
def variant_selectors(self) -> dict[str, int]:
return dict(self.active_variants or {})
def inventory_root(self) -> str:
return str(self.inventory_dir).rstrip("/")
def _resolve_variant_payloads(
*,
roles_dir: str,
include: Sequence[str],
active_variants: Mapping[str, int],
) -> dict[str, Any]:
"""Return ``{app_id: variant_payload}`` for the requested round.
Apps without `meta/variants.yml` collapse to a single empty variant
in the loader, so this just picks variant 0 (= `meta/services.yml`
unchanged) for them. Out-of-range indices clamp to 0.
"""
variants_per_app = get_variants(roles_dir=roles_dir)
resolved: dict[str, Any] = {}
for app_id in include:
variant_list = variants_per_app.get(app_id) or [{}]
if not variant_list:
continue
index = active_variants.get(app_id, 0)
if not 0 <= index < len(variant_list):
index = 0
resolved[app_id] = variant_list[index]
return resolved
def _bake_overrides(
*,
base_overrides: Mapping[str, Any],
variant_payloads: Mapping[str, Any],
) -> dict[str, Any]:
"""Merge the per-app variant payloads under `applications:` into the
`--vars` JSON the init step will hand to `infinito create inventory`.
Caller-supplied `extra_vars` (already inside `base_overrides`) win
over the variant bake when the same key is set on both sides — that
matches the existing "user vars always win" rule."""
merged: dict[str, Any] = dict(base_overrides)
if not variant_payloads:
return merged
existing_apps = merged.get("applications")
if not isinstance(existing_apps, Mapping):
existing_apps = {}
apps: dict[str, Any] = {
app_id: payload for app_id, payload in variant_payloads.items()
}
# Caller-supplied `applications.*` entries deep-overlay the variant
# payload so overrides like `applications.web-app-foo.feature_flag` from
# `--vars` still take precedence.
for app_id, override in existing_apps.items():
base_payload = apps.get(app_id)
if isinstance(base_payload, Mapping) and isinstance(override, Mapping):
apps[app_id] = {**base_payload, **override}
else:
apps[app_id] = override
merged["applications"] = apps
return merged
def build_dev_inventory(compose: Compose, spec: DevInventorySpec) -> None:
"""Build the inventory folder described by `spec` and ensure its
vault password file exists. Two side-effects, no return value: after
this call, `spec.inventory_dir` contains a complete, variant-resolved
inventory.
"""
inv_root = spec.inventory_root()
variant_payloads = _resolve_variant_payloads(
roles_dir=str(compose.repo_root / "roles"),
include=spec.include,
active_variants=spec.variant_selectors(),
)
vars_payload = _bake_overrides(
base_overrides=spec.overrides(),
variant_payloads=variant_payloads,
)
cmd = [
"infinito",
"create",
"inventory",
inv_root,
"--host",
"localhost",
"--ssl-disabled",
"--vars-file",
DEV_INVENTORY_VARS_FILE,
"--vars",
json.dumps(vars_payload, sort_keys=True),
"--include",
",".join(spec.include),
]
if should_use_mirrors_on_ci():
mirrors_file = generate_ci_mirrors_file(compose, inventory_dir=inv_root)
cmd += ["--mirror", mirrors_file]
extra_env: dict[str, str] = {}
if spec.services_disabled:
extra_env["SERVICES_DISABLED"] = spec.services_disabled
compose.exec(
cmd,
check=True,
workdir="/opt/src/infinito",
extra_env=extra_env or None,
)
_ensure_vault_password_file(compose, inventory_dir=inv_root)
def filter_plan_to_variant(
plan: list[tuple[int, str, dict[str, int]]],
variant: int | None,
) -> list[tuple[int, str, dict[str, int]]]:
"""Pin a matrix plan to a single round when `variant` is set.
`variant=None` returns the plan unchanged (full-matrix mode). An
explicit variant index is matched against the plan's round indices;
if it's out of range, raises `ValueError` so callers can surface a
clean operator-facing error rather than silently doing nothing.
"""
if variant is None:
return plan
for entry in plan:
if entry[0] == variant:
return [entry]
available = sorted(idx for idx, _, _ in plan)
raise ValueError(f"variant {variant} out of range; available rounds: {available}")
def plan_dev_inventory_matrix(
*,
roles_dir: str,
include: Sequence[str],
base_inventory_dir: str,
) -> list[tuple[int, str, dict[str, int]]]:
"""Return ``[(round_index, inventory_dir, round_variants), ...]``.
`total_rounds = max(variant_count)` across the included apps. In each
round R, every app uses variant index R when its variant list is long
enough, otherwise variant 0 (the legacy `meta/services.yml` payload).
Inventory paths are suffixed with `-<round>` only when `total_rounds
> 1`, so single-variant deploys keep the historical unsuffixed path.
Pure function — does not touch disk.
"""
if not include:
raise ValueError("plan_dev_inventory_matrix: include must not be empty")
variants_per_app = get_variants(roles_dir=roles_dir)
variant_counts = {
app_id: max(1, len(variants_per_app.get(app_id) or [{}])) for app_id in include
}
total_rounds = max(variant_counts.values(), default=1)
base = str(base_inventory_dir).rstrip("/")
plan: list[tuple[int, str, dict[str, int]]] = []
for round_index in range(total_rounds):
round_variants = {
app_id: round_index if round_index < count else 0
for app_id, count in variant_counts.items()
}
inv_dir = f"{base}-{round_index}" if total_rounds > 1 else base
plan.append((round_index, inv_dir, round_variants))
return plan
def build_dev_inventory_matrix(
compose: Compose,
*,
base_inventory_dir: str,
include: Sequence[str],
storage_constrained: bool,
runtime: str,
extra_vars: Mapping[str, Any] | None = None,
services_disabled: str = "",
) -> list[tuple[int, str, dict[str, int]]]:
"""Build every folder in the matrix plan and return that plan so the
caller can iterate it (typically by handing each folder to the deploy
stage in turn)."""
plan = plan_dev_inventory_matrix(
roles_dir=str(compose.repo_root / "roles"),
include=include,
base_inventory_dir=base_inventory_dir,
)
for _round_index, inv_dir, round_variants in plan:
spec = DevInventorySpec(
inventory_dir=inv_dir,
include=tuple(include),
storage_constrained=storage_constrained,
runtime=runtime,
extra_vars=extra_vars,
services_disabled=services_disabled,
active_variants=round_variants,
)
build_dev_inventory(compose, spec)
return plan
def _ensure_vault_password_file(compose: Compose, *, inventory_dir: str) -> None:
inv_root = str(inventory_dir).rstrip("/")
pw_file = f"{inv_root}/.password"
compose.exec(
[
"sh",
"-lc",
f"mkdir -p {inv_root} && "
f"[ -f {pw_file} ] || "
f"printf '%s\n' 'ci-vault-password' > {pw_file}",
],
check=True,
)