* feat: add domain virtual responses with a compact editor and bounded edge matching
* fix: keep traffic working during rolling edge script updates
* refactor: centralize edge management transport
* chore: keep generated review artifacts out of git
* fix(dashboard): keep single apps in the service settings flow
* fix(web): refine Arabic landing and navigation
* fix(core): align host channel branding with dashboard copy
* fix(mail): complete local backups and portable server restores
* test(mail): report engine diagnostics on round-trip failure
* fix(mail): stop Postfix writers before container snapshots
* fix(mail): apply Postfix shutdown in the daemon action renderer
* test(mail): install GNU tar on the remote fixture host
* fix(build): keep environment values out of Dockerfiles and logs
* test(build): exercise secret transport in deployment CI
* test(build): align cloud fixture file and command identities
* fix(build): validate secrets before allocating build resources
* fix(mail): persist mail logs, fail2ban database and ignoreip across engine updates
Mail engine updates recreate the container (docker rm -f + docker run),
so anything not on a bind mount is lost: /var/log (mail.log, dovecot and
fail2ban logs), fail2ban's SQLite database, and operator ignoreip
customisations in /etc/fail2ban.
- Bind-mount /var/log and /var/lib/fail2ban from the host mail state dir.
Deliberately not /etc/fail2ban: a seed-once mount would hide later image
changes (same trap as #392's master.cf).
- Render operator ignoreip additions (OPENSHIP_MAIL_F2B_IGNOREIP, passed
through engine.env) into a jail.d override on every boot, as baked
jail.local value + operator value, so image defaults are never dropped.
- Run cron under supervisord so logrotate's daily job actually fires for
the now-persisted logs; logrotate is installed explicitly (only a
Recommends of rsyslog) and build-gated.
Fixes#1701
* Preserve existing mail state before replacement and repair supervised log rotation
---------
Co-authored-by: Hydra <abdelhamedmo59@gmail.com>
* Protect support intake and add mailbox sender blocking
* Regenerate support operator route reference
* Wire contact widget key into the website runtime
* Fix repeat contact challenges and classify private support blocks
* feat: add platform migrations and refresh the Cloud experience
* fix(ci): run MongoDB 8 imports on compatible runners
* docs(web): add Railway and Vercel migration pages
* feat(dashboard): add shared dotenv text editing and top add action
* fix(dashboard): wrap text editor actions on narrow screens
* perf: reuse Docker builders and parallelize independent Compose work
* test: exercise declared dependency gates on the isolated Docker host
* fix: preserve dependency ownership and native cache maintenance
* fix: drain dependency readiness failures through the shared task barrier
* test: align server recovery expectations with source scan recovery
* fix: retain released dependency providers during publication recovery
* Keep route verification and bound parallel deployment work safely
* Send scoped cancellation confirmations through the customer email queue
* Offer help without treating one canceled subscription as account churn
* Invite cancellation feedback and offer direct support
* Preserve native Docker dependency cache across source changes
* Keep runtime dependencies ahead of changing API source
* Split independent Docker regression suites across CI jobs
* Use fresh connections for routing configuration probes
* Show published deployments while final startup verification runs
* Complete verification translations and align test types
* fix(billing): keep overview usage and payments inside each plan
* test(deploy): align source recovery assertion with current main
* fix(billing): preserve Manage payment tab across organization switches
* fix(billing): match pending plans through existing Cloud references
* feat(billing): unify plan management and referral navigation
* docs(billing): describe plan management and align summary assertions
* test(billing): align named plan and accessible summary assertions
* fix(build): scale the managed build-cache budget with the disk
The managed project-cache budget was min(5 GiB, 10% of the Docker
filesystem). A project whose BuildKit cache outgrew 5 GiB was evicted by
the post-build prune after every build, so every deploy started cold.
- buildCacheKeepBytes is now 20% of the Docker filesystem, so the budget
follows the disk in both directions: disks under 50 GiB also move from
10% to 20%. This also applies to managed Cloud workers, which share
the collector.
- Below the 20% free-space GC target the idle budget shrinks to what
still fits above it (never below the old min(5 GiB, 10%)), so idle
caches give space back before rollback artifacts and running apps.
BuildKit's own 20% minFreeSpace still bounds a running worker.
- pruneManagedBuildCaches defaults to that budget instead of 5 GiB.
- The daily sweep marks its call as automaticRetention, so managed caches
use the same budget; keepStorageBytes still bounds native Docker cache.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* feat(servers): configurable build-cache budget per server
Adds an optional managed build-cache budget for remote deployment
servers, edited from a Build cache card on the server Overview tab and
through the existing server update API, CLI and MCP tool. Empty keeps the
automatic share of the Docker filesystem.
- servers.build_cache_budget_bytes (migration 0224), bounded 1 GiB to
64 TiB in the database and the API contract.
- Only remote deployment servers expose and accept it: the local row
builds through the host daemon's native cache and migration sources
never build, so updates there are refused; managed Cloud servers keep
rejecting every field except name. Changing it requires server admin,
since a large budget lets caches hold most of the disk.
- SSH Docker runtimes read the value through a new buildCacheBudgetBytes
connection option on every collection, so a change applies to the next
build without reconnecting. It sets the post-build prune budget,
BuildKit's maxUsedSpace and the daily sweep's managed budget, capped at
the filesystem minus the admission reserve and still reduced under
storage pressure; a failed read falls back to automatic.
- Updates are audited like gpuOverride and do not drop retained SSH
sessions.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* docs: regenerate the MCP and CLI references for the build-cache budget
The server update tool description and the `server configure` help now
mention the build-cache budget, so the generated references changed.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
* fix(migration): frame streamed rsync progress updates
* fix(migration): preserve exact progress with rsync byte units
---------
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Hydra <abdelhamedmo59@gmail.com>
The managed project-cache budget was min(5 GiB, 10% of the Docker
filesystem). A project whose BuildKit cache outgrew 5 GiB was evicted by
the post-build prune after every build, so every deploy started cold.
- buildCacheKeepBytes is now 20% of the Docker filesystem, so the budget
follows the disk in both directions: disks under 50 GiB also move from
10% to 20%. This also applies to managed Cloud workers, which share
the collector.
- Below the 20% free-space GC target the idle budget shrinks to what
still fits above it (never below the old min(5 GiB, 10%)), so idle
caches give space back before rollback artifacts and running apps.
BuildKit's own 20% minFreeSpace still bounds a running worker.
- pruneManagedBuildCaches defaults to that budget instead of 5 GiB.
- The daily sweep marks its call as automaticRetention, so managed caches
use the same budget; keepStorageBytes still bounds native Docker cache.
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
* fix(updates): make manual security advisories actionable
* fix(updates): classify conditional advisories as unverified reviews
* fix(updates): retain service state in grouped advisory reviews
* fix(github): fall back to a writable connected server's stored credential when no local server exists
Repository preparation borrowed a server's stored GitHub credential only
from the organization's canonical local server. Self-hosted setups whose
deploy target is a connected SSH server with its own stored PAT could
deploy but not scan the same private repo: with no server selected, the
fallback found no local server and preparation failed with
GITHUB_TOKEN_REQUIRED.
When no server is selected and the organization has no local server, use
the oldest connected deployment server the caller has write access to
that holds a stored token credential, the same set a real deploy can use.
Repository access, same-organization, server write permission, and
read-only local-read limits are unchanged. Explicit server selection is
unchanged.
Fixes#1630
* fix(github): require server selection for ambiguous repository credentials
* fix(dashboard): keep repository recovery server selection manual
---------
Co-authored-by: Hydra <abdelhamedmo59@gmail.com>
* feat(mail): expose fail2ban status, ban and unban via API and MCP
* feat(mail): manage Fail2ban bans from the existing Health tab
* fix(mail): preserve exact IP display in RTL ban reviews
---------
Co-authored-by: Hydra <abdelhamedmo59@gmail.com>
* fix(edge): serve a static site's 404.html for missing paths
* fix(routing): make static SPA fallback explicit and preserve defaults
* test(routing): update static fallback checks and run HTTP coverage
* fix(i18n): complete inbound alias guidance across locales
* test(monitoring): isolate the session in stream fixtures
---------
Co-authored-by: Hydra <abdelhamedmo59@gmail.com>
* Add OpenTelemetry Collector to the apps catalog
Adds otel-collector as a one-click app template (docker-compose kind,
category: analytics), alongside Grafana and the other observability
apps. Uses the upstream otel/opentelemetry-collector-contrib image
unmodified — its baked-in default config already stands up OTLP gRPC
(4317) and HTTP (4318) receivers plus a debug exporter, so no custom
command/config override is needed to get a usable collector running.
Also wires up brand-mark resolution:
- AppLogo.tsx: catalog logo "otel-collector" → simpleicons "opentelemetry"
- ServiceIcon.tsx: image name → same brand mark, for services created
from a raw compose file rather than the catalog (both
"opentelemetry-collector-contrib" and "otel-collector" match)
catalog.json regenerated from apps/catalog/*.json per the drift test in
apps/catalog.test.ts (gen-catalog.ts requires bun, unavailable in this
environment — reproduced its exact merge/sort logic with a plain node
script; verified the new entry lands alphabetically and the rest of the
file is a no-op diff).
* feat: add Beszel to app catalog
* feat(catalog): add Prometheus
One-click Prometheus (prom/prometheus:v3.15.1) for the monitoring stack
requested in #1432. Metrics and config live in named volumes, so scrape
targets added to /etc/prometheus/prometheus.yml survive restarts and
redeploys. Prometheus has no login, so the connection notes point users to
domain password protection.
Refs #1432
* fix(ui): add the Prometheus app logo
The dashboard's APP_LOGO map needs an entry for every catalog app
(app-logo.test.ts). Use the simpleicons prometheus mark, like Grafana.
Refs #1432
* feat(catalog): add Loki
One-click Loki (grafana/loki:3.7.9) for the monitoring stack requested in
#1432. Logs live in a named volume and are kept for 31 days through the
compactor's retention flags. The image is distroless, so there is no shell
healthcheck and the flags are passed with commandArgv. Loki has no login,
so the connection notes point users to password protection.
Refs #1432
* fix(web): ship the Loki logo on the website too
Vendored app logos must be byte-identical in the dashboard and website
public folders (scripts/app-marketplace.test.ts).
Refs #1432
* feat(catalog): add VictoriaMetrics
One-click VictoriaMetrics (victoriametrics/victoria-metrics:v1.154.1) for the
monitoring stack requested in #1432. It is protected by default: basic auth
with username admin and an auto-generated password, passed through
-envflag.enable so it never appears in the command line. Metrics are kept
for one year in a named volume; /health stays open for the healthcheck.
Refs #1432
* feat(catalog): add standalone pgAdmin4 app
Adds pgAdmin4 as its own one-click catalog app, separate from the existing
PostgreSQL + pgAdmin bundle (postgresql.json). This one ships with no
database of its own — it's for administering any reachable PostgreSQL
server, including another Openship-deployed database's private connection,
a managed cloud database, or anything else on the network.
Reuses the exact service shape already proven by postgresql.json's pgadmin
service (same image/version, port, healthcheck): dpage/pgadmin4:9.18,
PGADMIN_DEFAULT_EMAIL + PGADMIN_DEFAULT_PASSWORD (the officially documented,
fully deterministic setup — no random/log-only credential to work around),
surfaced directly in the connection card.
Also maps the pgAdmin4 brand mark in packages/ui/src/AppLogo.tsx: simpleicons
carries no pgAdmin glyph (cdn.simpleicons.org/pgadmin4 404s), so it resolves
to the project's GitHub org avatar, the same pattern already used for
Ollama, vLLM, and Miniflux.
Verified against the real image (dpage/pgadmin4:9.18): booted it standalone
with the same env vars, confirmed the /misc/ping healthcheck responds once
gunicorn finishes initializing (needs the ~60s startPeriod already used by
postgresql.json).
Verified: bun scripts/gen-catalog.ts + bunx vitest run src/apps/catalog.test.ts (46/46 passing),
bun run --cwd packages/ui lint (tsc clean).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
* feat(catalog): add Listmonk app
Adds Listmonk (newsletter/mailing-list manager) as a one-click catalog app:
a private Postgres instance plus the Listmonk service, non-interactive
install + Super Admin creation on first boot via the officially documented
LISTMONK_ADMIN_USER/LISTMONK_ADMIN_PASSWORD env vars, surfaced directly in
the connection card.
commandArgv runs "./listmonk --install --idempotent --yes && exec
./listmonk" — install is a no-op on redeploy (idempotent flag), so the
same command is safe on every boot, no separate prepare step needed.
Verified against the real image (listmonk/listmonk:v6.2.0), not just docs:
- Ran the exact install+serve command against a live Postgres, confirmed
the superadmin user is created non-interactively ("creating superadmin
user 'admin'" in the log).
- The image's default app.address is localhost:9000 (unreachable from
outside the container) — set LISTMONK_app__address=0.0.0.0:9000, confirmed
it then actually listens and responds.
- Logged in via POST /admin/login with the exact generated password,
confirmed an authenticated session (302 + session cookie, 200 on an
authenticated /api/settings call).
- Confirmed /health returns 200 for the healthcheck.
- bun scripts/gen-catalog.ts + bunx vitest run src/apps/catalog.test.ts: 46/46 passing
(added minEngine: 0.6.6, required by the schema whenever commandArgv is used).
Continues M3 (Tier B: app + one DB), first of this batch.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
* fix(apps): declare a host port so the app can't collide with others
Without a ports entry the host port equals the container port, so a second
app on the same port fails with 'port is already allocated'.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
* fix(apps): declare a host port so the app can't collide with others
Without a ports entry the host port equals the container port, so a second
app on the same port fails with 'port is already allocated'.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
* fix(apps): use a unique host port, not the container port
In port mode the host port is taken from the ports entry, so 'N:N' still
collides with whatever already holds that port (80, 3000, ...). Match the
other catalog apps and publish on a distinct host port.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
* fix(apps): use a unique host port, not the container port
In port mode the host port is taken from the ports entry, so 'N:N' still
collides with whatever already holds that port (80, 3000, ...). Match the
other catalog apps and publish on a distinct host port.
Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>
* feat(catalog): add Appwrite and Alertmanager with verified community recipes
---------
Co-authored-by: Giovanni Martins <giovannidealmeidamartins@gmail.com>
Co-authored-by: anshori <anshoriacc@gmail.com>
Co-authored-by: Asad Khan <55049186+AsadKhanAb17@users.noreply.github.com>
Co-authored-by: alkomy <alkomy@icloud.com>
Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>