test_backup_nfs_excludes.py (4fcf097e5b) computed the repository root
with Path(__file__).resolve().parents[6], which
test_project_root_import rejects outside a package __init__.py
(quality-high pass 66). The package now exports PROJECT_ROOT exactly as
its sibling files/__init__.py does, and the test imports it.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The NFS backup rsynced the whole export and honoured only one exclude, the
shared backup root. Volumes declared `backup: false` were copied anyway and
then pulled a second time by remote-2-local; in the swarm DR drill of
Magento#0 (run 40625679270) docs_sites alone (a full repo clone plus its
prepared and work trees) took about 15.5 GiB per copy and filled the
runner. The volume backup already skipped these through the backup_volume
lookup; only the NFS path ignored the declaration.
- utils/storage/nfs.py: volume_dir() names the state directory a volume
binds to (pinned name or key), shared with compose_volumes;
backup_disabled_dirs() returns every NFS-backed `backup: false` volume
dir that no backed-up volume also binds to.
- The role-local lookup backup_nfs_excludes feeds them into
BKP_NFS_2_LOCAL_EXCLUDES next to the backup root; script.sh takes any
number of excludes. Compose installs still pass only the backup root.
- recover.py protects every live state directory the snapshot does not
hold, so `rsync --delete` keeps the excluded volumes (model weights,
the docs tree, caches) and volumes created after the snapshot instead
of wiping them.
- The swarm DR drill restores BACKUP_NFS_VOLUME, the first NFS volume
not declared `backup: false`; PRIMARY_NFS_VOLUME stays for the drain
tests. BookWyrm now drills bookwyrm_media; decidim, fusiondirectory,
meta, coturn, svc-registry-cache and svc-registry-docker have nothing
backed up on NFS and skip the drill.
Measured: compose_volumes renders identically before and after for 291
roles across compose/swarm x local/nfs (1940 of 1940); the 15 excluded
directories are exactly the backup-disabled device dirs; the systemd unit
renders shell-safe on ansible-core 2.20 and 2.21 and stays byte-identical
in compose and swarm-local. A new unit test runs the real script.sh over
two generations and the real recover.py back into the export, and fails
without the restore protection.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
320929e17c read the tempdir compose.yml back with Path.read_text, which
test_no_uncached_file_reads rejects (quality-high pass 64).
utils.cache.yaml.load_yaml keys its cache on path, mtime_ns and size, so
the file sync() rewrites in place is read fresh; the label assertions
would fail on a stale read.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The autoformat step of `make quality-high` reflowed three lines of
320929e17c that exceeded the line length; tokens are unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Swarm deploys with `docker stack deploy --resolve-image never`, so a task
only rolls when its spec changes. Since fdf1a33c5f BookWyrm, Decidim and
the Bluesky broker build under a constant tag (bookwyrm_custom:3.11-slim
and friends): a ref bump rebuilds the image through the build-args digest
and pushes it, but the image string in compose.yml stays the same and the
running tasks keep the old image.
registry_sync now stamps every service that runs a locally built image,
siblings sharing the tag included, with the container label
`infinito.image-id=<local image ID>`. A rebuild changes the ID, the label
changes the ContainerSpec, and swarm rolls the tasks; an unchanged image
leaves the label and the deploy untouched. List and mapping label forms
are both kept intact.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
baudolo stops every container on a volume before the authoritative copy
unless its image is listed via --images-no-stop-required. Leaving the
marker out silently picked "stop": the Stalwart webmail start that the
backup cycle triggered got SIGKILLed mid `composer update` (CI run
40625679270), and 71 services that mount a named volume had never
decided either way.
- The marker is now `backup.stop_required` (true = stop first, false =
copy live), replacing the double-negated `no_stop_required`; the 50
existing declarations are inverted unchanged. backup_image takes the
marker and the wanted value explicitly:
lookup('backup_image', 'stop_required', false) still feeds
--images-no-stop-required, lookup('backup_image', 'disabled', true)
the skip list.
- The 71 undecided services carry researched values (repo files plus
upstream backup guidance: GitLab, Jenkins, Elastic, RabbitMQ, Typesense,
Qdrant, OpenLDAP, Prometheus, Moodle, ClamAV, Alertmanager, ...): engine
files, git repositories, SQLite, mnesia and multi-file state stop; image
code, write-once blob stores with metadata in the database and
atomically rewritten files copy live. Stalwart's main service stays
stopped until /var/lib/stalwart is verified to hold only logs.
- Re-obtainable data is no longer backed up at all (`backup: false` on
the volume): model weights, ClamAV signatures, SearXNG config/cache,
transient RabbitMQ queues, Funkwhale's Typesense index, the meta clone,
the agent-broker socket dir, coturn's unused data dir, and volumes no
image reads (Decidim's /app paths outside /code, BookWyrm's /app/data
which v0.9.3 never reads, FusionDirectory's measured-empty dir).
- The backup lookup only matches what actually runs once the role
declares it: Bluesky's broker (`custom: bluesky_broker`) and Decidim
(`custom: true`) declare their custom build, BookWyrm's worker/migrate,
EspoCRM's daemon/websocket, Moodle's cron and OpenProject's proxy use
their owner's image, and the templates drop their render-time
overrides. Decidim deploys decidim_custom:0.31.6 and BookWyrm
bookwyrm_custom:3.11-slim from now on; a BookWyrm ref bump still
rebuilds through the build-args digest.
- test_backup_stop_decision_declared requires the boolean on every
service with its own image that mounts a backed-up named volume;
test_stop_required replaces the old integrity test.
Measured against HEAD: 43 -> 78 services copied live, none moved the
other way; 14 -> 30 volumes skipped.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
roles/web-app-nextcloud/tasks/01_fullstack.yml: both sides added a
replicated-template task at the same spot; store-i18n's Talk signaling
backend list and the updater's Janus config now stay as two tasks with
their own conditions.
roles/web-app-nextcloud/templates/compose.yml.j2: the updater rendered
the talk service's container_volumes a second time inside the direct-media
branch, next to store-i18n's unconditional call, which doubles the
volumes/configs keys once direct media is on. The janus_conf mount carries
its own direct_media `when` in meta/volumes.yml, so the single call
renders it; the duplicate is gone.
DEPLOYMENT_MODE == 'compose' became IS_COMPOSE_MODE | bool in
web-app-nextcloud/meta/services.yml, web-svc-coturn/vars/main.yml and
tasks/stages/03_destructor.yml, as store-i18n's deployment-mode lint
requires; the updater's coturn, Talk and decommission tests now pass
IS_COMPOSE_MODE alongside DEPLOYMENT_MODE. test_talk_direct_media.py
expects store-i18n's talk command line and mount position, and the
"host networking ignores direct media" check moved to the config level,
where direct_media already requires network_mode bridge.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
CodeQL flagged five notes on PR #1215: four test modules imported unittest both as a module and through from-import, and sys-front-inj-matomo's __init__ exported PROJECT_ROOT without declaring it, so the import from test_site_cache.py read as unused. The modules now bind mock through import unittest.mock as mock, and the package declares __all__, the same fixes main carries from 3610e7b9ef and 6e776a5780. The same pattern is fixed in the new Talk direct-media test and the decommission integration package, and the coturn test package drops an __init__ export nothing imports.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
janus_conf is mounted only when direct media is on, which requires compose mode, and compose ignores a config mode outside swarm and warns about it on every deploy. The file keeps the 0644 it is rendered with on the host.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
web-opt-rdr-domains renders a `return 301` vhost for each network sibling
of a redirect source (network_redirects), but the health expectations only
forced the clearnet source to [301]. A handover app's onion sibling kept
its [302] handover expectation, so Bluesky's disabled view domain failed
"view.bluesky.<onion>: ERROR: Expected [302]. Got 301." and
hlth-webserver aborted the deploy. Every expected domain whose clearnet
sibling is a redirect source now expects [301].
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Removing a role from the inventory left its compose stack running under /opt/compose, holding ports, memory and a stale instance directory, until someone stopped it by hand.
On a full compose deploy with MODE_CLEANUP (no --id), the destructor now takes every directory below DIR_COMPOSITIONS that is named after a known role, carries the full instance layout and is owned by no active role (inventory-wide groups, group_names, the dependency closure and the instances this play prepared), stops it with compose down --remove-orphans without -v, verifies through the project and working_dir labels that no container is left, and only then deletes the directory; named volumes stay. Hand-made, partial or backup directories and instances a non-compose container is named after are left in place and listed. The step is the last block of the destructor, with the timer unlock in its always section, so a failure cannot leave timers locked.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The aio-talk image runs Janus with its stock janus.jcfg: no rtp_port_range and no nat_1_1_mapping, so Janus offered only its bridge address and every stream of a call went through coturn as relay. Calls above about six participants stuttered and failed.
When talk runs bridged next to standalone coturn in compose mode on a host that owns networks.internet.ip4, a rendered janus.jcfg is mounted over the stock one with rtp_port_range 61000-61999, nat_1_1_mapping = ip4 and keep_private_host, and that UDP range is published on ip4. TURN stays the fallback. Before rendering, the deploy reads the image's launcher and stock janus.jcfg and stops when Janus no longer reads the mounted path or the stock settings drifted from the template. Variant 0 publishes 64 ports so dev and CI exercise the path; the call capacity spec asserts a non-relay selected pair on ip4 in that range. With the flag off every rendered file is byte-identical to before.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
With host networking and no --relay-ip, coturn put every non-loopback IPv4 of the host into its relay list, Docker bridge gateways included, and handed IPv6 and bridge clients relays on those addresses while advertising the public ip4, so their media went nowhere. It also listened on every bridge address.
In compose mode with host networking, coturn now listens on networks.internet.ip4, 127.0.0.1 for the healthcheck and networks.internet.ip6 when set, and relays on ip4. It pins only when the host owns ip4, and owns ip6 too or leaves it unset; behind 1:1 NAT, without the ip binary or with a foreign ip6 it renders the previous command unchanged, because coturn exits when it cannot bind a listening address. A unit test renders the real template and vars for every cell.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
16 conflicts, resolved per hunk:
- sys-stk-front-proxy: tasks/main.yml keeps the network-sibling loop;
the single stack_host_contents write of http.conf plus vhost and the
front_proxy_vhost_changed gate move into tasks/vhost.yml, with the
http.conf decided per sibling domain so an onion vhost gets no HTTPS
redirect.
- sys-front-tls: 01_http.yml is gone as the updater intends; the
cert-renew include in main.yml gates on the vhost domain's TLS.
- sys-svc-compose: the updater's resolve_host_cid without the remote
echo and the prepared pending flag in up.yml, written with
IS_SWARM_MODE and the repeated-when block grouping.
- web-app-nextcloud: the updater's post-upgrade gate and pending-only
plugin configuration, which already carry the pinned-plugin handling.
- web-app-xwiki: one XWIKI_CORE_VERSION for the readiness probe and
style.css.j2; OIDC/LDAP follow xwiki_extensions_ready at bootstrap,
the S3 blob store stays off there as TestBlobstoreRenderOrder demands.
- roles_with_service: onion swap and dashboard card key combined, the
canonical domain is the canonical URL's host in both branches.
- sys-ctl-hlth-csp: --clearnet-timeout and --concurrency together;
run_checker forwards its options to build_docker_cmd, keeping the
script under the 500-line limit.
- web-app-matrix: Element branding and web_client_location side by side.
- web-app-stalwart/README.md: the generated Cosmos section is dropped.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
csp-checker 3.2.0 checks every URL in its own browser context and takes --concurrency for a worker pool. The role pins 3.2.0, sets HEALTH_CSP_CONCURRENCY to 4, and the systemd unit hands it through script.py to the container command ahead of the -- URL separator, for the clearnet and the onion batch alike. TimeoutStartSec keeps its sequential sizing.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
test_include_uniform_apply_keyword flagged legacy_checkout.yml after 11c2d8dcd3 moved its guard onto the include: both tasks repeated become: true. The include in 03_launcher.yml now applies become to the file and keeps the guard; the order test reads include_tasks in its mapping form too.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Each plugin ran its own occ config:app:set per configuration value right after its install or enable, and a redeploy rewrote every value. Install and enable now queue the plugin; the configure stage runs once for the queue: it loads every queued addon spec, reads occ config:list once, keeps only the values Nextcloud does not hold exactly (nextcloud_plugin_config_pending; typed, masked and structured values are always kept), and runs the hooks in queue order afterwards.
Combined with store-i18n's 7f72e53a58: the pending occ config:app:set calls go on the stdin of one container exec -i ... sh -s, so no value (Talk, OnlyOffice, whiteboard, BBB secrets) reaches a host argv; a failing call names its appid/configkey.
Replay harness old vs new: identical final state, updated keys and hook order in first-deploy, converged, drift and failure scenarios (failure stops at the same key); converged config:app:set calls 19 -> 3. Merge note: resolve 03_configure.yml with this side against 7f72e53a58.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Every app's compose pass ran three remote tasks before compose up: create the directories, compose ps for running containers, and stat the pending marker (53 apps x ~2.7 s in a production run). compose_instance_prepare does all three in one call from 02_stack_host_prep.yml, sharing the directory creation with directory_batch through utils/directory_batch.py; compose ps runs only in compose mode. up.yml evaluates the registered result, named compose_instance_before_files because it predates 04_files: the running check is skipped whenever the compose, env or override file changed in this run (those changes notify compose-up themselves).
Replay harness old vs new: rendered files and the compose-up, marker and swarm handlers identical in every scenario; the only difference is the skipped running check when only the override changed, where compose-up still runs. About 85 s per full production run.
Merge note: store-i18n's a4c7fda974 hoists the removed marker stat into a block; resolve up.yml with this side.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
DISCOURSE_REPOSITORY_DIR appended the repository URL to the services directory, so the launcher lived under services/https:/github.com/discourse/discourse_docker.git/. It now lives under services/discourse_docker. Before the clone, launcher/legacy_checkout.yml removes the URL-named tree (computed with relpath against the services directory) and prunes its emptied parents; a guard skips the cleanup when that path would escape the services directory or equal the new checkout. The cleanup must precede the clone: git otherwise resolves the old relative path as a URL and aborts with 'protocol ... is not supported'. Discourse data stays in the launcher's named volume and /var/discourse/shared; the first deploy rebuilds once because the bootstrap digest lived in the old containers/ directory.
relpath joins the builtin filters test_defined knows.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Every redeploy ran occ maintenance:repair --include-expensive (43.5 s, maintenance mode on) and listed the appstore updates through Tor (49.7 s) although the version had not changed. 03_upgrade.yml now reads occ status and files/mimetype_version after occ upgrade; nextcloud_needs_post_upgrade decides whether the pass is due (version changed in this deploy, or the expensive repair has not run for the running version). A due deploy includes 03_upgrade/01_post_upgrade.yml with today's app update and repair; an unchanged one includes 03_upgrade/02_cached_updates.yml, which updates only the unpinned enabled apps whose release stored by the updatenotification background job is newer (nextcloud_cached_app_updates, PHP version_compare semantics). NEXTCLOUD_PLUGINS_PINNED is the single source of the pinned check.
Verified against echoserver's live occ output (status, marker, updatenotification cache, app:list, with debug log lines): unchanged redeploy is not due and finds no update. Scenario harness 24/24 incl. negative controls on the old file. About 86 s per unchanged redeploy.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Conflict in roles/sys-stk-front-proxy/tasks/main.yml: store-i18n moved
the per-vhost tasks into the network-sibling loop of tasks/vhost.yml, so
main.yml keeps HEAD and the deferred probe lands in vhost.yml instead:
every sibling domain is queued in front_proxy_deferred_probes, and the
inline probe of an unchanged vhost is gone in favour of the batched one.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Each vhost used to run its own sequential reachability probe. The role now queues the vhost and deferred_probe.yml checks the whole queue with one uri_probe_batch call after the shared-service preload and after the server roles. A vhost answering like a stale config restarts OpenResty once; one still unserved afterwards fails the deploy by name. uri_probe_batch takes follow_redirects explicitly so the probe sees the redirect status itself.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Conflicts resolved:
- sys-front-inj-matomo: inject.yml keeps the store-i18n split; the cached
failed site list load (5b7284c4a4) is ported into resolve_site.yml next
to its token guard, and test_site_cache.py reads that file.
- web-app-bluesky: platform_administrator from store-i18n with the
prefix-free handle and its i8- reserved variant from feature/updater;
the broker CONFIG keeps siteTitle, storePath and chooseHandlePath.
- web-app-stalwart README: the managed Quick Setup section is dropped,
the docs role renders it.
- README.md: the role overview is regenerated from the merged tree.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
5b7284c4a4 named its reset task with 81 characters, one over the limit of
test_name_emoji_and_length, and 754d35c4de built the path to
roles/web-app-keycloak/vars/main.yml with a pathlib chain that test_file_spot
requires to come from ROLE_FILE_VARS_MAIN. Both commits went in without the
lint suite. The task keeps its behaviour under a 70-character name, the test
reads the same file through the constant. make test-lint: 577 passed.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
keycloak_kcadm_update built `cat <<'JSON' | kcadm update|create ... -f -`
with the JSON payload inside the command it hands to sh -c, and the
reCAPTCHA task did the same in its shell script. Client secrets, the SMTP
password, the LDAP bindCredential and the reCAPTCHA secret therefore sat
in a host process argument (#1213).
The library now sends the payload as subprocess input, also on its lock
and dead-container retries, and the reCAPTCHA task uses the shell
module's stdin. Every caller's kcadm_exec already runs `container exec -i`.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
While the private CA is missing from the prometheus stack, a self-signed
target has to use http_ready_private_ca. The blackbox job chose that
module once per application from lookup('tls', app, 'mode'), which aligns
to one sibling domain. For web-app-docs that sibling was the onion domain
(TLS off), so its clearnet https target kept the verifying http_ready and
failed with "x509: certificate signed by unknown authority" (CI run
40131272595: n8n#1, and hidden behind the first failure in API#1 and
Mattermost#1).
Every served domain now gets its own target group whose module follows
lookup('tls', probe_domain, 'mode'). Series labels are unchanged.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
SAML clients were created only by the realm import, and the import runs only
for a missing realm. echoserver keeps its legacy realm cymais.cloud and ran
SuiteCRM behind oauth2-proxy before v16, so the SuiteCRM SAML client
https://<suitecrm>/saml/metadata never existed there: the SAML hop showed
Keycloak's Invalid Request page and Keycloak logged LOGIN_ERROR
error=client_not_found reason=Cannot_match_source_hash.
The existing-realm reconcile now also loops over the clientId of every SAML
client in the rendered realm dictionary, the same source the import uses,
and creates or updates it through kcadm like the application-declared OIDC
clients. The same path already converged the stalwart-webmail and
stalwart-webui clients in echoserver full-10.
Same change as 2b4db90576 on feature/store-i18n, so a later merge of the two
lines resolves without a conflict, plus a unit test that pins the SAML client
selection and its place in the reconcile loop.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
On a host whose Matomo was not deployed yet, every vhost injected before
web-app-matomo asked SitesManager.getSitesWithAtLeastViewAccess again and
waited out uri_retry (30 retries x 2 s): 6 vhosts, 84 to 98 s each, 8.7
minutes in galaxyserver full-2. matomo_site_cache only kept a loaded list.
inject.yml now caches every executed load: the list on success, none on a
failed request or an error mapping. A cache of none skips the cached-id step
and getSitesIdFromSiteUrl, so the existing skip task turns the injection off
for the vhost as before, without the retry wait. web-app-matomo resets the
cache of every play host (delegate_facts, swarm workers included) after
07_config and before 08_reinject_consumers, so the re-injection and every
later vhost load the list from the now running Matomo.
Evidence: the real inject.yml with the real uri_retry against a closed port
and a stub Matomo, 11 vhosts in 4 scenarios: the first probe took 68.6 s, the
next two 0.4 and 0.3 s with matomo off; after the reset a known site came
from the cache, an unknown one went getSitesIdFromSiteUrl -> addSite and was
appended, a later vhost of that domain hit the cache. HEAD inject.yml took
68 s per vhost with the same outcome. The swarm reset reached a second host.
ansible-lint passes; tests/unit/python/roles/sys-front-inj-matomo pins the
gates and the reset order.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
One string that every engine refused or mangled raised out of the batch route's gather, so the gateway answered the whole batch with a 422 and the client lost the other nineteen translations. Gateway.translate_each now owns the bounded fan-out; with partial=true it answers (None, None) for a string that ran into a GatewayError and keeps the rest. Non-gateway exceptions still raise, and a NoBackendError on every string still fails the batch as the outage it is.
The i18n client asks for partial batches and counts a null answer as a refused entry with its reason, so the re-ask and damage checks continue per string. A caller that does not send the flag gets the old behaviour.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The reconcile task retried every failure thirty times ten seconds apart, so a permission the key lacks surfaced only after five minutes: rdr-3 spent 331 s on a 403 that no retry could change. The reconciler now ends with exit code 3 when Flowise answers 401 or 403, and the task stops retrying on that code while its failed_when still fails the run under MCP_RECONCILE_STRICT.
Only 401 and 403 count as final: authorize may answer a provider that is still starting with another 4xx, and that wait is what the retries are for. Every exit that reports an HTTP status goes through fail(), which keeps the message on the exception, so the existing tests read it unchanged.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
test_mint_mcp_api_key.py computed PROJECT_ROOT itself from Path(__file__).parents[6], which test_project_root_import forbids outside an __init__.py. The files package now exports PROJECT_ROOT the way its reconcile subpackage does, and the test imports it.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The reconciler prunes the managed fixture chatflows of providers it no longer probes, but the scoped API key it gets was minted without chatflows:delete, so the first prune ended in 403 Forbidden after thirty retries; on echoserver rdr-3 stopped on 'FAILED deleting stale infinito:fixture:web-app-prometheus'. Adding the permission alone would not have reached that host, because api_key() reused an existing key unchanged.
KEY_PERMISSIONS now carries chatflows:delete, and api_key() reuses the managed key only while its permissions equal KEY_PERMISSIONS; otherwise it deletes the key through DELETE /api/v1/apikey/<id> and mints a new one, which the reconciler receives in the same run. The script gains its mirrored unit test, so its mirrored-unit-test exemption goes.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Every addon ran through the install, enable and configure stages on its own, and each stage asked occ for the plugin's state again. The role now reads occ app:list --output=json once, and the new nextcloud_plugin_plan filter replays the addon order (an enabled entry switches its incompatible_plugins off and itself on, a disabled entry switches itself off, the later entry wins) and splits the app ids into disable, install, enable and configure. One app:disable call takes every id that has to go off; the install, enable and configure includes loop only over the ids that need that stage, so an already enabled plugin skips straight to its configuration. A plugin shipped as a pinned archive still goes through install on every run.
On echoserver the plugin phase took 608 s in full-9 for 51 addons; a repeat deploy is derived at about 195 s instead of about 550 s.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
quality-high's autoformat pass rewrote three files that 9301220d05,
49aa920034 and c78d5f906c committed without the hooks. ruff drops three
`noqa: E402` markers as unused (RUF100): E402 does not fire on an import
that follows sys.path.insert. The formatter wraps the long
KEYCLOAK_RBAC_GROUP_LOGIC_DIGEST line and the OpenProject admin runner's
`sh -c` line; the double-quoted scalar and the `>-` block both fold the
break back into a space, and a parse against HEAD gives equal values.
On merge: feature/store-i18n 2bcf59109e carries the same four Keycloak
hunks byte for byte; they merge without a conflict.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
ruff drops three noqa: E402 markers as unused (RUF100): E402 does not fire on an import that follows sys.path.insert. The formatter also wraps the long KEYCLOAK_RBAC_GROUP_LOGIC_DIGEST line; the double-quoted scalar folds the break back into a space.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The emailVerified update looped over every non-reserved user, uploaded a
shell script per user and ran kcadm with a JVM login each time, and it
always reported changed: 7 users took 23.8 s on echoserver.
email_verified.py uses the shared keycloak_admin client through
tasks/utils/rest_script.yml: it looks each user up by exact username,
skips users already verified, and writes emailVerified with a GET plus a
full-representation PUT, as kcadm update did. The task reports changed
only when a user was updated, and a failed lookup now fails the task
instead of being treated as a missing user. The old lookup piped
kcadm -q username=X into head -n1, which could have hit a longer username
sharing the prefix.
Evidence (local compose, slot 104): summary updated=0 unchanged=8
missing=0 failed=0 in 0.60 s, task ok, staging dir removed.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The per-app RBAC groups and LDAP group mappers were reconciled by four
kcadm shell scripts that started a JVM per call and re-synced every app
on every deploy: 582 s for the mappers plus 324 s for the orphan purge on
echoserver (39 apps).
rbac_reconcile.py now talks to the admin REST API from inside the
Keycloak network namespace (nsenter --net) with the admin credentials
read from /proc/<pid>/environ, so nothing secret reaches argv or output.
It drops the shared ldap-roles mapper, purges root groups of undeployed
apps, ensures each app group and mapper, and syncs a mapper only when the
group is new, the mapper config drifted, or the fingerprint stored on the
group (mapper payload, app rbac and domains, group-logic digest) differs.
The per-app payload no longer carries groups.path.group: Keycloak 26.8.1
does not store that key, so a payload with it never matched the stored
mapper and every deploy re-synced every app.
The admin client lives in keycloak_admin.py; tasks/utils/rest_script.yml
stages it with the script in a 0700 temp dir on KEYCLOAK_HOST_NODE, runs
it and removes the dir in an always block.
Evidence (local compose, slot 104): matrix iteration green in both
variants and passes; update runs show synced=1 in 1.1 s, then
synced=0 unchanged=1 in 0.81 s through the final delivery path, no
staging dir left behind.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
On a multi-network host one gateway serves the clearnet domain and its onion sibling, but allowedOrigins carried only the canonical URL. The Tor pass opens http://claw.ai.<onion>, so the gateway closed every Control UI websocket with 4008 before the pairing step and the UI reported origin-not-allowed (142 onion origins, 120 closes in the Funkwhale#1 rescue log of run 40042690056).
OPENCLAW_CONTROL_UI_ORIGINS takes lookup('network_urls', OPENCLAW_URL): one URL per network sibling with that sibling's scheme, canonical first, and the URL alone on a single-network host.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A chat engine prefixed the Vietnamese catalog with its own preamble ("...Vietnamese:" followed by a blank line) and wrote Urdu in Devanagari. Neither tripped the damage predicate: vi is Latin script, so the script checks were blind, and untranslated only measures the Latin share, so a non-Latin language answered in another non-Latin script passed.
added_lines refuses a translation that breaks more lines than its source, and the gateway's _mangled carries the same check so such an answer falls through to the next engine instead of being cached. wrong_script refuses a non-Latin language when less than half of the letters it writes in another language's endonym script belong to its own; the script per language is derived from the endonym in meta/languages.yml, and letters are counted over runs of letters plus combining marks so Indic and vocalised Arabic words stay whole.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The parallel-sessions test asserted that more than one of four sessions
appended something, which a loaded host where one worker drains the
shared iterator fails with nothing broken. The sessions now meet at a
barrier before their first append, so a serialised pool still fails the
assertion while a concurrent one always passes.
(cherry picked from commit 657015852a)
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
The documentation index renders README.md, whose badges load from img.shields.io, and one static build is served on every network the node runs, so the onion docs page requested a clearnet third-party host and network-leak.spec.js failed on the docs+tor pass (fork run 40042690056, jobs Moodle#1, Collabora#1, Listmonk#1 and the swarm API#1 job, log line 37249).
A Sphinx extension now downloads every remote image of an HTML build into the doctree cache, names it by the hash of its URL with the suffix its Content-Type names, and points the image at the local copy so Sphinx ships it under _images. An image that cannot be fetched gives way to its alt text and keeps its link, and the gettext build, which runs with -W, is left alone. Visitors' browsers no longer contact img.shields.io at all.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
With the asset and design services off (dashboard variant 1) the logo falls back to service_provider.platform.logo, which without web-svc-asset is the public https://file.infinito.nexus/assets. network_url leaves a host the node does not serve unchanged and port-ui embeds it as is, so the onion dashboard requested clearnet https://file.infinito.nexus/assets/img/logo.png and network-leak.spec.js failed in four jobs of fork run 40042690056 (Gitea#1, OpenProject#1, Mastodon#1, XWiki#1). Pure tor mode loads the same URL; its leak spec only runs with more than one pass.
csp.yml now lists the asset service URL under img-src. build_csp_header does not read img-src, so the header is unchanged; external_asset_origins does, which puts the host on the web-svc-mirror allowlist and into the onion body filter's rewrite to http://mirror.<onion>/file.infinito.nexus, the path every other third-party asset of an onion page takes. A deployed web-svc-asset resolves to the node's own file domain, which external_asset_origins skips. test_branding_mirror_origin.py covers both cases.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
With the Keycloak OIDC directory active, Stalwart refuses every account create with 'forbidden: Cannot set credentials for accounts in an external directory', and full-4 failed the new batch assert on all accounts. The create request now lists the accounts in the same call and takes each id from the listing of the provisioned domain. Under SSO a forbidden create is tolerated for people, who may not exist before their first login; every bot still needs an id because its app password is addressed by account.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
#1180 gave web-app-matrix a canonical `jitsi` domain for its own Jitsi
and listed it under `services.jitsi.domains`. When Jitsi is shared or
disabled nothing serves that domain, and the blackbox healthz job of
web-app-prometheus probed every canonical domain, so the probe of
https://jitsi.matrix.<domain>/healthz/ready failed with "tls:
unrecognized name" and `prometheus scrape: every consumer role reports
up=1 and passes its probe` went red (main run 40017166046: Matrix#2,
Nextcloud#12).
web_health_expectations already skipped such keys. The rule now lives
in utils/domains/unserved.py: a key claimed by a service entity that is
disabled, or enabled but shared, is not served by that app. The health
filter uses it unchanged, and prometheus.yml.j2 filters its targets
through the new `served_domains` filter, which also drops the
`<key>_onion` mirror of a skipped key.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
