radarr reached its download clients as nzbget:6789 and qbittorrent:8112 and its indexers
through jackett:9117 or indexers.zurag.be - HAL container names and a public route,
typed into its database by hand. Nothing on the mesh answers those names.
It now requires nzbget-api, qbittorrent-api and jackett-api. A run-once step
(downloads/, declared last, restart-on its bindings, credentials and settings) writes
host, port, TLS, base path, user and credential into radarr through radarr's own API:
- A download client is the mesh's when its name is downloads.<provision>.name and its
kind the provider's; it is registered when missing. A jackett feed is the mesh's when
its host is the bound one, one the step bound before, or one in
downloads.jackett-api.adopt-hosts; the jackett indexers in
downloads.jackett-api.indexers are registered when missing. Every other entry is left
alone, and nothing is ever deleted.
- Only connection fields, only when they differ. The stored password is masked, so the
app tests the entry with the credential it holds; only if that fails is the delivered
one written. Categories, priorities and "enabled" are never touched.
- Each credential is tried against its provider first. A minted value (nothing accepted
yet) is never written; the step exits 1 naming the exact secret accept.
The step is byte-identical in sonarr, radarr, lidarr and bookshelf (the same Servarr
API, v3 or v1): each module builds from its own directory, so each carries a copy, and
test/downloads.test.ts fails if a sibling's copy differs.
Verified: strict typecheck, the Dockerfile build, 14 unit tests; and the compiled step
against fresh pinned sonarr/radarr/lidarr/bookshelf with throwaway nzbget, qBittorrent
and jackett - minted credentials refused with nothing written, accepted ones registered
and tested by the app, a migration-shaped radarr repointed, reruns unchanged.
The manifest named /services/radarr/config and /var/lib/mesh/radarr/route.json, host
paths ADR 0112 takes out of definitions. The config dir is now a pathless ${dir:config}
(0700, 1000:1000) mounted into the server and, read-only, into the runtime that reads the
API key from config.xml; the route binding lives in a placed state dir, as jackett and
searxng do. /var/lib/mesh/radarr stays for the broker secret.
The image is pinned to 6.4.4.10685-ls317, the digest ace runs today. The old pin
(6.3.0.10514-ls314) was older than the running version, and Radarr migrates its database
forward on start, so pointing the older build at ace's data is not safe.
The container mount points stay /movies and /downloads: Radarr's database stores its root
folder and the download clients' reported paths under exactly those names, and there are
no remote path mappings to absorb a change. The generic access paths stay; ace's
(/storage/media/movies, /storage/downloads) and its media owner 1001:2000 wait on hq 153.
Based on feat/servarr-api-provision (#156), which makes radarr provide radarr-api.
Verified: catalogue key tests with MESH_CATALOGUE set (73 manifests parsed, not skipped);
a throwaway container of the pinned image on a fresh 0700 1000:1000 config dir answers
/ping, serves the v3 API with the key it generated and refuses a wrong one (401), accepts
/movies as a writable root folder; the runtime's client discovers the key from that
config.xml and reads the queue.
A consumer on another machine (ombi first; jackett, bazarr and home-assistant
later) reached these by container name on HAL's shared network, which the mesh
does not have. Each app now provides <app>-api at mesh scope and serves the
software port, so the mesh tells a consumer where it is and redirects the
port to where the machine published it.
Named per app, not one servarr-api: a requirement is matched by name and
answered by exactly one provider per node, so a consumer cannot require one
name from three providers - and ombi's code is written against each app's
own API version (ADR 0027).
No grants and no provisioner: a Servarr instance has one API key, which the
mesh cannot mint. The operator accepts it as the pair credential for each
consumer (ADR 0092).
A host-network sidecar reaches its service over the machine's loopback, and
the mesh publishes that service on a machine port it assigns (ADR 0038) —
so dialling the software's port reaches whatever else holds it. On ace,
searxng's sidecar dialled 127.0.0.1:8080 and got unifi's inform port. The
same shape in bazarr, bookshelf, lidarr, nzbget, qbittorrent, radarr and
sonarr; each now asks with ${port:N} (hq 088). Found in review of ace's
module preparation.
75 endpoints across 50 modules, named from what each one is for rather than by a
rule: mail's seven protocol ports are smtp, imaps, submission and the rest; unifi's
nine are inform, stun, discovery, the two portal ports and syslog; minio's two are
s3 and console; the resolver's two are dns-udp and dns-tcp.
And 35 route contributions name the endpoint they serve instead of repeating its
port. A route and a listen both carried a port and nothing said they were the same
thing; now one of them does. gitea's path-level deny rule names neither, because it
is a rule about a name rather than an endpoint.
novox/hq ADR 0138. The words shipped a release ahead in mesh-controller #138 and
#139, and the control plane running today is built from that merge — checked before
this was written, because an unknown manifest key is refused and a catalogue using
one against an older control plane would stop resolving.
Every module named its events the way the old bus spelled a routing key —
`module.<module>.<verb>`. Design 29 says a module names an event locally and the
mesh works out where it lands, so all 37 were stale against a rule already
decided. On the new bus that derives into a namespace belonging to a module
called "module", so no cross-module subscription in the mesh matched anything:
nothing failed, nothing reacted (novox/hq 04-ISSUES/127).
36 manifests converted, and 43 files of module code with them. The code mattered
as much as the manifests: the runtime builds the subject from what `emit()` is
handed, so a converted manifest with unconverted code would have had the
permission and the subject disagree.
Three things the new check found on the way:
- `photos` emitted an event its manifest never declared, which the new bus refuses
outright. Declared.
- `showcase` waited for an event nothing emits, so its demo could never be
triggered — only `showcase` may publish under its own name. It emits both halves
now.
- `distribution` declared an event named after a different module. It emits
`image.pushed` under its own name. An event about a *role* belongs on the seat,
where the name outlives whoever holds it, but the sdk has no way to publish on a
seat yet, so that stays recorded rather than declared.
The audit logger's "everything" pattern is `**` rather than the old bus's `#`.
The 21 media/home modules, the SaaS tool modules (cloudflare-dns,
confluence, gitlab, jira), model-usage, and the model-access trio get
the same Dockerfile + build section as batch 1. Scheduled-only modules
(anthropic-consumer, anthropic-manager, openai-consumer) deliberately
declare no MESH_TOOL_MODULES — every container of theirs names its
command. mosquitto's run-once bootstrap container builds from the same
artifact. Modules with third-party deps (model-usage: pg;
anthropic-manager: tweetnacl) install them beside their compiled code.
Also fixes cloudflare-dns's package.json, unparseable since its
description lost a closing quote.
Deliberately still without build sections: builder and mesh-controller
(the foundation builds them by its own path), distribution (provides
the artifact store — building it through itself is refused by design),
route-proxy (cross-repo build context, deferred), and the
upstream-image-only modules, which have no code to build.
Piece A + B of ADR 0056, completing the internal-CA work in ff01ada.
Selectable issuer. acme-ca is now a role two providers can satisfy: step-ca
(internal CA) or the new public-acme (a fact-only module, no container/listen)
that serves Let's Encrypt production. A mesh assigns one or the other to satisfy
route-proxy's `requires: acme-ca`.
One directory shape for both. A provider serves the ACME directory's parts the
way the mesh already models any reachable service -- an address (`at`), a `port`
and a `path` -- and route-proxy composes `https://<at>:<port><path>`. step-ca
lets the mesh fill `at` (its node) and `port` (its single listen) and serves only
`path`; public-acme, not being a mesh service, serves all three (overriding `at`
with the public host). Same composition either way.
Empty root means the system trust store. Both providers serve `root`: step-ca
the operator root PEM (settled per mesh), public-acme an empty string. route-proxy
writes it to the CA bundle file unconditionally; the binary now reads an empty
bundle as "the root is already trusted by the OS" and falls back to system roots
(examples/route-proxy/main.go, committed on the mesh-control ADR-0056 branch).
step-ca inits from the operator's root. The operator's root cert, root key and
root-key password are mounted at the smallstep entrypoint's default init paths
(/run/secrets/root_ca.crt, root_ca_key, root_ca_key_password) with the matching
DOCKER_STEPCA_INIT_*_FILE vars, so `step ca init` adopts the operator's root
instead of self-generating one -- the CA that signs is the CA route-proxy trusts.
Route names are labels, not FQDNs. Every routed module now contributes a `label`
(the leftmost subdomain) instead of a full public hostname; the node's public
domain composes the name. Apex (novox.be) is left as a full name -- composeName
has no empty-label/apex convention yet (mesh-control follow-up).
Claude-Session: https://claude.ai/code/session_01LrgweAeERJYBg88c5cKDzF
Add a `route` contribution (requires/contributes/binds) to every web app so
each gets a Host-routed public name via route-proxy, mirroring the
de-spiegel/only-office pattern:
- novox: gitea, keycloak, nextcloud, umami, invoicing, verdaccio, registry,
and mailu (single mail.novox.be -> 7080; admin/webmail/api ride that port).
- ace: grafana, sonarr, radarr, lidarr, bazarr, ombi, tautulli, jackett,
nodered, searxng, home-assistant, bookshelf, baserow.
Split photos so its three sites each get a name: photos keeps server +
admin-client (photos.novox.be), and new photos-eef (eef.novox.be) and
photos-filip (filip.novox.be) modules carry the client sites.
The production FQDN stays the literal default; a per-node .incus name is a
settings override applied where the mesh runs, not a manifest hardcoding.
Claude-Session: https://claude.ai/code/session_01LrgweAeERJYBg88c5cKDzF
The radarr/sonarr/lidarr/bookshelf events entrypoints built their client
with XClient.fromEnv() at import time, which throws when no API key/URL is
available yet — crash-looping the runtime container. The tools already
guard this; the events entrypoint did not.
Mirror the tools' guard: build the client in a try/catch, and only start
the poll loop when it succeeds. When it fails, log one line and stay idle
until a key is available. Behaviour when configured is unchanged.
Claude-Session: https://claude.ai/code/session_01LrgweAeERJYBg88c5cKDzF
04-ISSUES/036: eight media modules each declared the shared library and
download directories under /services/media/* as their own `directory`
resources. Six of them owning one path is the collision the resolver
refuses — so the stack's only sensible assignment, all on one machine
sharing one filesystem, would be refused the first time two landed
together.
The media library is the operator's, owned by no module (novox/hq
ADR 0051). Move every /services/media/* path from an owned `directory`
resource to an `accesses` entry: the mesh mounts it and owns nothing —
does not create, chown, reconcile or remove it — and several modules may
access one path with no conflict. Each module's own config and mesh-state
directories stay owned resources.
Modes are least-privilege: plex reads the libraries it streams; the
managers and download clients get read-write on what they import and
write; bazarr writes subtitles into the libraries (read-write) and only
reads the download spool. The container volume mounts are unchanged.
Claude-Session: https://claude.ai/code/session_01LrgweAeERJYBg88c5cKDzF
Mirrors plex: a broker-bound runtime that serves the module's tools, discovering
the app's API key from its own config.xml under a read-only config-dir mount, URL
defaulting to the server on the node. Proven in the mesh-lab: assigned-sonarr green
(key detected, tools served under the scoped account, no live Sonarr needed).
Claude-Session: https://claude.ai/code/session_01LrgweAeERJYBg88c5cKDzF
The review found broker own-secret paths drifting: mostly /var/lib/<module>/
broker, but grafana/redis/icecast/nextcloud carried a '-module' suffix to dodge
a collision with the service's own /var/lib/<name> data, and photos sat under
/etc. Normalized to one collision-free namespace a service never owns:
/var/lib/mesh/<module>/broker, with a mesh-state directory resource for the
parent, across 24 modules. audit-logger is grandfathered (lab-proven, referenced
by the assigned test, and it has no service to collide with). All 33 manifests
parse; no client hardcoded a path, so nothing in code moved.
Claude-Session: https://claude.ai/code/session_01LrgweAeERJYBg88c5cKDzF
The Servarr apps (TV, movies), each self-contained from hal's shared arr
client. Tools: status, library, search, queue, calendar. Events by polling the
download queue: a new item emits module.<app>.<thing>.grabbed, an item that
leaves as completed emits module.<app>.download.completed — the exact key plex
consumes to rescan. A grab that left as failed/warning is not reported as a
completion. Both typecheck; manifests parse.
Each module becomes modules/<name>/ holding module.json, with room for
the rest of what a module is — its tools, health checks, provisioning,
lifecycle — which the conversion from hal still has to bring across.
The flat <name>.json was only the resource-declaration half.
Claude-Session: https://claude.ai/code/session_01LrgweAeERJYBg88c5cKDzF
jschoubben