The nfs backup rsyncs the whole export, and the shared backup root
lives inside it (every node's DIR_BACKUPS is the mounted state dir), so
each generation embedded a copy of all previous backups and grew
quadratically; the giant listings also flushed real errors out of
captured journals. The unit now passes the backup root relative to the
export as an anchored rsync exclude, and recover.py protects the live
root from --delete while never replaying embedded stale copies from old
generations (DirectoryRecovery gains rsync_extra_args for that).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
meta/tests.yml becomes the SPOT for per-role test metadata: the
test-deploy discovery skip list moves there (top-level skip:) from the
primary service entity in meta/services.yml, next to the existing
cli.timeout. get_role_skip reads the new location; the drv lint,
--skip-mode help and unit fixtures follow.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
svc-runner's CLI test runs a full nested reinstall deploy and needs
75+ minutes, so the uniform 3600s timeout killed it deterministically
in both modes. A new role_tests lookup reads roles/<id>/meta/tests.yml
(cli.timeout, scaled through the timeout lookup) with the existing
default as fallback, keeping the short hang-leash for every other CLI
test. svc-runner gets 4h and pins the nested deploy to variant 0; the
second matrix round only redeployed cdn+dashboard and added no
assertion value.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The swarm labs run NFS server and clients on one shared kernel; with
hard mounts a writeback stall wedges containers in D-state and takes
the whole DinD node with it (the bbb/matomo and nextcloud wedges).
dev/act already mount soft for exactly this reason; github gets the
same lab profile. Production (host) keeps hard mounts.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Replace the find_dock_val_by_bkp_entr filter with three role-local
lookups that read the live merged applications config:
backup_database_containers (backup.database_routine -> service names),
backup_image (backup.no_stop_required/disabled -> exact deployed image
refs via the image lookup, swarm registry prefix and custom builds
included), and backup_hard_restart_projects
(backup.project_hard_restart -> compose project names, declared on
mailu instead of hardcoded). Pin baudolo to 3.0.0, whose --images-*
flags match .Config.Image exactly and whose hard-restart flag is
--hard-restart-projects.
The restore drill now starts all compose projects in parallel per
retry round; the previous serial alphabetical loop deadlocked when a
project's migrate service waited on a datastore project that sorted
later (mastodon -> postgres), killing the CLI test at the 3600s
timeout. Volume restore rsyncs into the mountpoint directly, dropping
the rsync helper image remnants from the test env.
BREAKING CHANGE: deploys require backup-docker-to-local 3.0.0.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The custom flag moves from 33 hardcoded template kwargs into the merged
service config: 'custom: true' names the shared <entity>_custom build,
a string names a second distinct '<string>_custom' image inside one app
(nextcloud whiteboard). The image lookup reads the declaration as a
fallback, so config consumers (e.g. backup image matching) resolve the
exact deployed reference from one declaration point. nextcloud cron now
runs the nextcloud_custom build like the app service instead of the
upstream image.
A new services lint couples the declaration to reality in both
directions: custom requires a role Dockerfile, and a role Dockerfile
requires a custom declaration. The eight legacy roles deploying local
builds under upstream tags or ad-hoc image= overrides are ratcheted via
'# nocheck: dockerfile-custom'.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
New 'image' lookup is the SPOT for resolving <registry>/<image>:<version>
(swarm registry prefix in swarm mode, bare in compose; custom= accepts a
string base for a second distinct custom image inside one app).
container_image becomes a thin wrapper that forwards terms+kwargs and adds
the compose image: "<ref>" line, so all existing template call sites keep
their output byte-identical. The lookup-usage integration test now counts
lookup_loader.get() delegation as production usage.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The OpenTalk controller v0.34.1 verifies its bucket via S3 ListBuckets
membership, but SeaweedFS hides every bucket from an unconfigured
identity (confirmed identical on 4.38 and 4.39), and the embedded sidecar
started with no s3.config, so the controller's signed request got an
empty bucket list and crash-looped on 'bucket missing' although the
bucket existed. The sidecar now runs s3.configure to register the
consumer's access key with Read/Write/List/Tagging on its bucket, the
same grant the central provider gets from its mounted s3.json.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
lookup('canonical_url', app_id[, key][, consumer=app]) resolves the web base
URL for an app or one of its named canonicals (element/synapse/filer),
following the onion-injected merged-domain map: .onion targets resolve to
http (Tor provides transport security), everything else follows TLS_ENABLED.
The consumer= kwarg returns '' unless the consuming app has the target's
service binding enabled, replacing the recurring
"(lookup(...) if lookup('config', app, 'services.X.enabled') else '')"
template branching.
- utils/cache/domains.py: inject the node onion into dict-form
(named-canonical) domain declarations, not just list/str forms
- keycloak redirect_uris: onion redirect URIs are http
- 13 playwright.env.j2: manual scheme branching -> canonical_url(consumer=)
- lint: forbid manual '://' concatenation and to_onion_url in role templates
- unit tests: 10 cases (primary/named x clearnet/onion, consumer gating)
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Replace scripts/system/diagnostics/rescue.sh with utils/diagnostics/
container.py: one workflow call on failure captures the current level
(containers with logs+inspect, swarm services, journal, host resources,
in-play role dumps) and recurses itself through every DiD boundary via
self-copy, from the runner down to the innermost runtime (RESCUE_MAX_DEPTH).
compose and swarm now share an identical failure step; the swarm per-node
collector wrapper is gone. All 17 script-only Ansible rescue: blocks are
deleted and their bare block: wrappers inlined (plus a wordpress leftover);
the three justified blocks (postgres pg_hba, xwiki extensions, keycloak
cross-node recovery) keep their evidence tasks and now end in a fail task.
Two lints enforce the state: rescue: is forbidden without a
'# nocheck: rescue-<reason>' first-task marker, and a block without
rescue/always/when or another scoping key must be inlined.
sys-service-terminator hoists become to the include apply. Also move
routine/08_cleanup.sh to utils/clean/teardown.sh and topology/ to
utils/topology/ so routine/ holds only the sequenced flow.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
svc-bkp-volume-2-local now declares secrets_backup (enabled+shared true):
host secrets are captured unconditionally wherever container backups run.
svc-bkp-remote-2-local declares local-2-device gated on group membership
(operator flag, no variants) so the pull host syncs to the encrypted
device when the role group is assigned; local-2-device registers as a
shared provider and gains the 01_core run-once schema. local-2-device
recover.py learns --device-target: snapshots live under <mount>/<target>,
so recovering without the configured target subpath found nothing.
secrets-2-local memory shrinks to 128m/256m (a oneshot rsync) to keep
variant stacks under the host memory budget.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Add routine/backup/: base.sh drives a marker seeded into the live NFS
volume and the manager secrets through the whole chain (volume+secrets+nfs
backup units -> remote-2-local pull -> local-2-device LUKS 'USB') and back
via every recover.py (device -> local root -> nfs export, docker volume and
host secrets), asserting each marker survives. Each per-host step is one
in-node routine run over docker exec. extend_inventory maps the volume/
secrets/nfs backup roles onto their hosts so the units install; write_extras
generates the pull keypair; the INFINITO_SWARM_BACKUP_* keys are added to
default.env + the passthrough handler.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
New svc-bkp-secrets-2-local role: a differential rsync --link-dest snapshot
of the host-generated secret material that lives in no docker volume, NFS
export or operator inventory - infinito secrets + tokens, the self-signed
CA, the Let's Encrypt account/certs, the ACME DNS credentials - plus the
node identity (ssh host keys + machine-id). files/recover.py restores each
present subtree to its fixed system path and gates the node-identity
restore behind --restore-node-identity. Wires the systemd unit + nightly
schedule (06_daemons, 07_schedule). Unit tests cover the domain/target map
and the main() subtree-selection + node-identity gating.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Add utils/recovery/base.py DirectoryRecovery: a service-backed pre-recover
(start the role's backup unit first) + rsync -a --delete mirror. Each
svc-bkp-* role gets a files/recover.py subclass (volume resolves the docker
mountpoint, nfs mirrors the export, local-2-device opens the newest LUKS
snapshot across machine hashes) or a documented one-way opt-out
(remote-2-local recover.py.nocheck). A lint guard enforces every svc-bkp-*
role ships a recover.py and a ## Recover README section; docs/architecture/
backup.md documents the end-to-end schema. Unit tests cover the base flow,
the per-role class logic and every recover.py main() CLI wiring.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Agent sandboxes run in a network namespace with an empty /proc/net/route
while /sys/class/net still exposes the host NICs, so dotenv generated the
static 1500 fallback although the host uplink is 1400 and every TLS
handshake from the DinD stack blackholed. Take the minimum MTU across UP
physical interfaces (device symlink present) before resorting to the
static default.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Two wrapper behaviours move onto single sources of truth:
Exec timeout: SYS_TIMEOUT_CONTAINER_EXEC (07_schedule.yml, default 7d so a
hard kill never beats a production migration; the dev/CI inventory
tightens it to 60min) reaches every task host via the play environment as
CONTAINER_EXEC_TIMEOUT. The wrapper prefixes only the exec subcommand with
coreutils timeout --kill-after=15. Ansible task timeouts are SIGALRM-based
and slept through the last wedged docker socket for 4.5 hours; the OS
level kill still fires there and returns 124, which until/failed_when and
the rescue diagnostics react to. Unset or 0 keeps todays behaviour.
CA container paths: the bind-mount targets for the injected root CA and
wrapper script lived as string literals in container.py, compose_ca.py
and the filter plugin. They now live once in the CA_TRUST group_vars SPOT
(inject_cert_container, inject_wrapper_container) and reach container.py
as env vars (play environment plus CA_TRUST_SYSTEMD_ENV) and compose_ca.py
as CLI arguments built by the compose_ca_inject_cmd lookup. A caller that
provides the host CA env but not the container paths is half-migrated and
fails loudly instead of silently dropping CA injection; a host without any
CA env keeps the soft fallback. All tests read the paths from the SPOT,
and a regression test asserts the runtime scripts carry no literal copy.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The three identical Exception-labelled comments over cache_base._reset()
restate what the call does; the reset itself is unchanged.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Run the python suites through pytest-xdist when pytest is importable
(id-level collection parity with unittest discover verified per suite;
serial unittest stays the fallback), with -rP so warnings printed by
passing tests, like the update annotations, still reach the log. run.sh
self-provisions the dev extras through scripts/install/dev-extras.sh,
which stamps inside the venv and skips read-only venvs; make onboard
pre-warms the same path in the running container.
Hoist per-test cache resets and roles scans to setUpClass in the lookup
and resolvable suites (a per-test fingerprint-memo reset closes an
id()-reuse cache-poisoning flake the hoist exposed), inject generation
names into the nfs backup script test instead of sleeping, and narrow
ansible-lint scope to real ansible content. ansible-lint additionally
sits behind a content fingerprint gate that skips the run when nothing
it walks has changed; without a usable git (container bind mounts) the
gate disables itself and the full run remains.
Drop the import-linter layer and ignore for the deleted deps module;
the contract now locks compose -> common -> env. Measured on the host:
unit 101s to 20s, integration 90s to 24s, lint tests 43s to 19s,
repeated ansible-lint 53s to 0.4s.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Rename the role, its test tree, the primary entity key
(volume-2-local), the derived vars (BKP_VOLUME_2_LOCAL_*,
SYS_SERVICE_BACKUP_VOLUME_2_LOCAL), the run-once fact, and every
group_names consumer flag, include path, doc, and lint reference. The
container_backup service key and the on-disk backup-docker-to-local
tree keep their names: both are contracts with existing consumer
entries, snapshot trees, and the baudolo pull tooling.
Operators must rename the svc-bkp-container-2-local inventory group and
disable the previous versioned units on already-provisioned hosts; the
pre-deploy snapshot glob derives its base from the new role name and
will not find them.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add a systemd-driven rsync --link-dest backup of the NFS export into the
baudolo-compatible local snapshot tree, plus a pre-deploy snapshot hook in
the constructor. A missing export fails the nightly unit loudly, a failed
rsync removes its incomplete generation so it cannot poison the next
--link-dest, and rsync exit 24 from the live export is tolerated.
Both pre-deploy snapshot scripts (container and nfs) now resolve the
previous deploy's unit through a version glob, so the snapshot still runs
after a project version bump. The user-backup ssh wrapper whitelists the
new backup-nfs-to-local type for downstream pulls. The nfs-server role
declares the backup as a dynamic consumer with true/false pinning
variants; svc-bkp-container-2-local drops its variants.yml because
non-subject closure members resolve foreign round indices against it and
leak its companion pins into other rounds.
Unit tests execute the real script against tempdirs and prove hardlink
deduplication by inode, history isolation, delete semantics, loud failure
on a missing source, and cleanup of incomplete generations.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The nfs SPOT loaders pulled PyYAML into the .env generation, which bootstraps fresh hosts BEFORE python dependencies exist: the import crashed cli.meta.env, load.sh produced an empty environment and make install-system-python died on the required-var guard (environment/manjaro and pkgmgr/centos CI jobs). The loaders now line-parse the services.yml SPOTs via the stdlib-backed cached reader and the mapping path constant. The client's protocol version field becomes nfs_version, since services.yml reserves version for image tags (schema lint), and the image discovery casts versions defensively.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
storage.nfs shrinks to the deployment-specific server address: export_base moves into svc-storage-nfs-server's services.yml (the server dictates where exports root) and the mount protocol version into svc-storage-nfs-client's services.yml (the client owns how it mounts). Both are read through utils.storage.nfs loaders by the nfs lookups, the env handler and 00_topology, eliminating the unrendered-variables class for these values; the group_vars assert shrinks accordingly and the swarm test extras stop restating role defaults.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Registering svc-bkp-container-2-local as the container_backup provider made the constructor's service-loader preload it in the server bucket, ahead of mailu (web-app bucket) - but its alarm chain (sys-ctl-alm-email -> sys-svc-mail) hard-asserts the mailu no-reply token, and run_after cannot legally cross buckets. A provider can now declare preload: false: it stays resolvable in the registry (consumer flags, closures, variants) while ordered_primary_service_entries skips it, leaving the existing constructor group pass (which runs after the full preload) as its load path.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Lookup plugins receive ansible variables unrendered, so routing the state subdir through a group_vars jinja value handed state_path the literal '{{ lookup(...) }}' string and the controller mount asked the server for /srv/nfs/{{ lookup('nfs_state_subdir') }} (Connection refused, nextcloud swarm roundtrip). Both nfs lookups now import STATE_SUBDIR directly; the intermediate lookup, the group_vars indirection and the unit-test fixture var that masked the gap are gone.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
The controller crash-looped on a missing S3 bucket and the recorder on an unprovisioned OIDC client while the swarm converge gate blocked before any post-flush task could create either. The embedded seaweedfs sidecar now bootstraps its consumer bucket itself at startup (new seaweedfs_sidecar_script filter: entrypoint-wrapped server, bounded readiness poll, idempotent weed shell bucket create, TERM/INT forwarded to the server), and the Keycloak clients are provisioned before the handler flush. Post-flush bucket tasks stay as an idempotent backstop; the obsolete restart tasks are dropped. Live-validated against chrislusf/seaweedfs:4.38: bucket created, stop behavior identical to the stock entrypoint.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Rename the meta/services.yml key across all roles, the roles_with_default_placement lookup (now roles_with_placement), the get_role_default_placement / iter_roles_with_default_placement helpers, and every consumer (swarm closure, compose_volumes filter, nfs_prep, container deploy template, constructor stage, extend_inventory) plus their tests and docs.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
make autoformat normalised inline-comment spacing (` #` -> ` #`) in role meta/services.yml + meta/variants.yml and re-wrapped a few Python files (ruff line-length). Pure formatting — verified via git diff -w (whitespace-only) and diff inspection (ruff line-wraps only, no logic change).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
New to_onion_url filter swaps a clearnet host suffix for the tor_node and forces http (no-op without tor_node or off-primary hosts). body_filter.lua uses it in onionize() so that .onion requests receive the onion CDN origin for injected head/body snippets (e.g. universal-logout.js), which the Tor browser cannot reach on clearnet. Covered by 7 unit tests.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add the global TIMEOUT_FACTOR and a shared timeouts.js helper: resolveTimeout(base) scales by TIMEOUT_FACTOR and, over a .onion target, an extra multiplier, so Tor's per-request latency no longer trips hard-coded timeouts. Migrate every raw timeout in the specs (timeout:/timeout=/setTimeout/waitForTimeout) to resolveTimeout, stage the helper in all three staging paths, and add test_no_raw_timeouts to forbid raw numeric timeouts. Clearnet behaviour is unchanged (factor 1: resolveTimeout(N) === N).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
web_health_expectations swaps (exclusive) or extends (dual) each services.tor.enabled app's clearnet domains with their onion domains via _inject_onion_domains, so the probe checks exactly what the deploy serves; script.py probes .onion over http (plaintext-only) with a longer Tor timeout. Adds onion unit tests for both.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add a `stack` column to the complexity matrix: true when a role renders
its own container stack (ships templates/*compose*.yml.j2), false for
host-only roles (backup crons, wireguard, swapfile), the swarm
infrastructure roles themselves, and central-webserver content
providers. Exposed as a sort key and filter field.
Make `swarm` imply `stack`: a role with no compose template of its own
is never a swarm stack-deploy target, so swarm is forced false when
stack is false. Because scripts/meta/resolve/apps.sh selects the swarm
deploy jobs via `--filter "swarm == true"`, this drops 17 stackless
roles (106 -> 89) from the swarm job list; all keep compose testing and
their swarm overlay coverage still runs through the webserver/proxy
role, so no real deploy coverage is lost.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
A curl against an accepted-but-silent peer hangs forever: task retries
never fire because the first attempt never returns, and the CI job dies
at the 6h runner cut (proven by the mailu admin-API hang that took the
suitecrm/peertube swarm legs down). The new `curl` filter renders the
hardened base command ("{{ 30 | curl }}" -> curl -s --connect-timeout 5
--max-time 30); scaling composes with lookup('timeout').
test_curl_via_filter enforces it: bare curl invocations in role tasks,
templates and shipped scripts fail the lint with a pointer to the
filter. Plain .sh files, which cannot render Jinja, pass by pinning
--max-time inline; Docker healthchecks are exempt (their timeout field
caps them); `# nocheck: curl-timeout` opts out (catalogued in
suppression.md).
Convert all 34 existing invocations: mailu (replacing the short-lived
MAILU_CURL var), elasticsearch, checkmk, yourls (300s for the
schema-creating installer POST), the telegram alert script, and the
plain scripts of jellyfin, svc-runner, apt_repo, moodle (600s tarball)
and the mattermost provisioner. Mailu's user/token tasks additionally
carry task-level timeouts so a wedged container exec cannot stall the
play.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Add registry.gitlab.com to the mirrorable registries so discovery
yields the GitLab CNG images to the version updater and mirror sync.
Seed the generic v2 tag scan with the OCI last cursor for v-prefixed
pins and raise the page size to 1000: CNG repos carry tens of
thousands of commit-sha tags that sort before the release tags, so an
unseeded scan exhausted the page cap without reaching any semver tag.
Pin every gitlab service version in meta/services.yml and drop the
GITLAB_VERSION template override so services.yml is the single source
for per-service versions.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The freeswitch build fetched libopusenc from launchpad.net, whose
librarian backend rate-limits CI (429s across 11 attempts); three more
build-time fetches carried the same risk. The maintainer publishes an
alangecker/* image for every buildable service, so compose_mods drops
build: wherever the service carries an image and compose pulls instead;
a guard task fails loudly if a future upstream service still builds.
Pin ref to the v3.0.23 release branch (the packaging repo has no tags)
and clone it without submodules via the new compose_repository_recursive
toggle: submodules only feed build contexts and DEV_MODE mounts, and
generate-compose falls back to the committed repos/tags file.
Declare the pulled images in meta/services.yml (updater-managed SPOT,
mirror pre-pull) and the named data volumes in meta/volumes.yml; the
override renders both through the standard container_image /
container_volumes / compose_volumes lookups with the service_name
pattern. Compose merges volumes by target, so the override's named
volumes verifiably replace the clone-local bind mounts (docker compose
config proven). This also heals a split-brain mount: apps-akka,
periodic and prometheus-exporter previously kept bind mounts while
their peers used the named volumes.
Teach container_image and compose_volumes to honour compose_mode_force
(the SPOT the other mode-aware lookups already read): bbb deploys as a
node-local compose stack in swarm, so it must get hub image refs and
plain local volumes there, not registry-prefixed refs and NFS
driver_opts. The unpinned-engine-volume lint learns the same two safe
conditions. Read the embedded Collabora port from web-svc-collabora's
services SPOT instead of hardcoding 9980.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Deploy GitLab from the official CNG CE images as six services: webservice (puma) and sidekiq scale to the topology default, workhorse is the sole HTTP entry, gitaly and a one-shot migrations job are pinned singletons, shell runs gitlab-sshd. The Omnibus container, its Dockerfile and the GITLAB_OMNIBUS_* config groups are removed. All rails/workhorse/shell/gitaly configs and shared secrets are pre-rendered by Ansible and mounted read-only; eleven credentials land in meta/schema.yml including an rsa_pem_2048 OIDC signing key (new value_generator algorithm). Postgres, redis and S3 object storage ride the existing lookups; OIDC, SMTP, the public SSH port and the root password are preserved. compose_volumes gains a per-volume nfs:false opt-out so the gitaly repositories volume stays node-local in swarm (upstream forbids git repositories on NFS); nfs_prep skips volumes without driver_opts accordingly. The Nextcloud gitlab integration execs the CNG rails runner. ruby_single_line dies with its only consumer; ruby_dq keeps test coverage. Ports (8181/8080/2222/8075/3808/9122) and service names resolve from single SPOT declarations in meta/services.yml via vars constants; the version pin sits on the webservice image pair so the updater can bump it. Live-validated in compose up to all six services healthy and migrations completed; fresh-install-only semantics for existing Omnibus data are documented.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Workers failed 'docker build: no Dockerfile': the build context, compose file, .env and runner binary were staged on the stack host only while every node deploys its own node-local runner stack. Build and push the image once on the manager (RUNNER_IMAGE_REF via the swarm registry, 04_image.yml) and let workers pull it; render compose.yml (template with lstrip_blocks, matching the replaced stack_host_template action), .env and the extracted run.sh on every node. Add a node_local kwarg to the compose_networks/container_networks/compose_restart/compose_only lookups so this node-local compose stack renders a bridge default network (workers cannot create overlays), keeps restart: unless-stopped (ephemeral runners died after one job) and keeps container_name under swarm; default is unchanged for every other role. Treat build/install.stamp as stale when the venv is missing and exclude ./build from the nested-src copy, closing the nested-e2e ruamel failure class. Regression tests for the node_local render paths; the network include tests now accept lookup kwargs.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Over Tor a page served on an .onion still advertised a clearnet-only CSP and oauth2-proxy config, so onion origins were blocked (dashboard could not frame auth.<onion>; oauth2-proxy rejected the redirect back to the onion app). build_csp_header now mirrors the clearnet primary to the node onion per token; the keycloak oauth2-proxy config derives whitelist_domains and cookie_domains from the app resolved domains. Exclusive (onion-only) apps get onion tokens only, primary apps get both, non-tor apps stay clearnet.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The !vault YAML constructor returned a plain str, so round-tripping a host_vars
file through utils.cache.yaml (e.g. apply_tor_node_onion writing
services.tor.node) re-emitted every credential as an untagged quoted string.
Ansible then never decrypted them and LDAP/DB admin binds failed with result 49.
Return a VaultScalar so SafeDumper re-emits the !vault literal block; add a
regression test asserting apply_tor_node_onion keeps existing vault credentials
tagged.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
On disk-constrained sync runners (act/CI) a mirrored external image was double-stored: once in the docker engine and once in the local registry. After pushing, docker rmi the upstream and prefixed tags so only the registry copy remains; swarm re-pulls from the registry at deploy time. Locally-built images are never removed (recovery and base images depend on them staying in the engine store). Also skip the re-pull when the prefixed image already has a registry manifest. The compose image rewrite and registry push are unchanged; rmi is best-effort.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Roles whose get_entity_name(application_id) resolves to an empty string (e.g. svc-runner) rendered 'name: ' with no value in the compose networks block, producing an invalid/empty network definition. Guard both the swarm overlay and the compose bridge emissions so the name line is only written when own_entity is non-empty; docker then auto-names the network. Add regression tests for both modes.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
make autoformat re-wrapped the hostname read_text over multiple lines and moved the # nocheck: cache-read to the closing .strip() line; test-lint stays green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Resolve services.<key>.<field> with consumer-override -> provider-native fallback (the provider being the role whose entity name is <key>, e.g. svc-net-tor for tor). Keeps provider defaults like tor.exclusive/tor.primary in one place while consumers stay free to override per app; consumed by utils.cache.domains for onion-domain injection.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Mint the node onion during provisioning and write its address straight into applications.svc-net-tor.services.tor.node, dropping the INFINITO_TOR_NODE_ONION env indirection: ensure_node_onion (was init_env) mints/reuses the .onion-identity key files + returns the address (no env write); identity_hs_dir/IDENTITY_DIRNAME move to utils.tor_onion so cli and utils share the path SPOT; provision.apply_tor_node_onion sets services.tor.node in host_vars when svc-net-tor is deployed (run by 'inventory provision' at deploy time, and the svc-net-tor role copies the same key files into the daemon); domains.py reads services.tor.node; nginx_vhosts reads the key file instead of os.environ; INFINITO_TOR_NODE_ONION removed from default.env/passthrough/compose + the inventory env-lookup. Bonus: the address now lives in the persistent key files, surviving .env regeneration by make test/autoformat.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Add a generic Docker Registry v2 client (WWW-Authenticate bearer-token
challenge) so tag listing and reachability work for any registry (Docker
Hub, GHCR, MCR, Quay, GitLab/opencode), and route the version updater
through it so quay.io and opencode.de images auto-update like the rest.
Add an external test that hard-fails when a pinned image:version tag is
unreachable (404) and warns on incomplete image/version pairs.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Checkmk's OMD site crashes at boot with fcntl.flock raising OSError
[Errno 9] Bad file descriptor: NFSv4 ignores the nolock mount option and
routes flock() to the server. Force flock() client-local on v4 so
single-writer apps (checkmk, openldap) boot; POSIX locks stay
server-side. NFSv3 keeps nolock (statd-free, flock already local).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Add a `tor_ports` lookup that collects every single-int `ports.public` entry
from the deployed roles' meta/services.yml and returns HiddenServicePort
mappings (`127.0.0.1:<port>`). UDP-only categories (relay, media) are excluded
since Tor forwards TCP streams only. Wire it into TOR_ONION_EXTRA_PORTS
(19_tor.yml) alongside the SSH port so mail (25/143/465/587/993), gitea/gitlab
SSH, ldaps, matrix federation and STUN/TURN control ports are reachable at
<node>.onion. Register Mailu's mail endpoints under ports.public and add the
matching PORT_BANDS entries.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
