Adds a Transfer action on every service page that moves an application,
compose or database to another server without S3: volumes, bind mounts
and deployment logs are streamed through the panel (ssh2/spawn pipe),
file mounts and Traefik config are recreated on the target, the source
is cleaned up and the service is deployed on the target. Failures before
cleanup roll back to the source server.
An Infisical provider is pinned to one non-recursive `secretPath`, so reading
two folders means two providers, two machine identities and two sets of
credentials to rotate. This lets a reference name the folder instead:
${{vault.my-provider.external/sentry:SENTRY_DSN}}
`<path>:<KEY>` mirrors the HashiCorp client in this directory, which already
documents that exact format. A relative path resolves against the provider's
`secretPath`, a leading slash is absolute, and a ref without a colon keeps its
current meaning — the whole ref is the secret name at the provider's own path.
A dot cannot be the separator here: Infisical accepts dots inside secret names
(`A.B.C` is a valid key), so `provider.a.b.C` cannot be split unambiguously
and would silently break anyone using such a name.
Refs are grouped by resolved path so each path is listed once, and the login
happens once per batch rather than once per path.
Tests cover the bare ref, relative and absolute paths, a provider at `/`,
grouping with a single login, the error naming the path, and a malformed ref.
Previously, upsertRecord matched only by type and name, silently
overwriting any existing record of the same type/name with different
content. This caused data loss when users had multiple records of the
same type (e.g. multiple TXT records for SPF + site verification).
Now upsertRecord also checks the record content/target before deciding
to update. If the content differs, a new record is created alongside
the existing one instead of replacing it.
Affected providers: Cloudflare, Porkbun, Infomaniak, OVH.
Tests updated accordingly.
A secret in Infisical can reference another one (`${env.folder.KEY}`), which
is the only way to keep a value in one place and read it from several folders.
The Infisical client fetches `/api/v3/secrets/raw` without
`expandSecretReferences`, and that endpoint leaves references untouched by
default. The referencing secret therefore arrives as the literal `${...}`
string, is written to the generated `.env`, and the deployment still reports
success — the service just receives a placeholder instead of its value.
This is easy to miss because the single-secret endpoint (`/raw/{name}`) does
expand by default, so the Infisical UI and API show the resolved value while
only Dokploy sees the literal.
Adds the parameter and the first tests for this client: the list request now
asserts the flag, a missing secret and an auth failure. Without the fix the
new test fails with `expected null to be 'true'`.
upsertRecord pulled every record of a zone just to find the one it was about to
write. The API can filter, so ask it to.
The match is still verified locally, because filter[source] matches substrings
rather than exact values. Verified against a live account:
filter[source]=autoconfig -> [autoconfig]
filter[source]=auto -> [autoconfig, autodiscover]
filter[source]=disco -> [autodiscover]
Letting the filter decide would therefore have upserted "auto" onto the existing
"autoconfig" record. The filter narrows what is transferred; the exact
comparison still decides. Covered by a test that fails if the first candidate is
trusted.
The apex was the other thing worth checking, since Infomaniak writes it as ".":
filter[source]=. does return the apex records, and an upsert on an existing apex
TXT returns its id rather than creating a second one.
Suggested by @narcisonunez on #5257.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
#5257 merged, so canary now carries the Infomaniak provider and its
0191_cool_christian_walker. This branch's 0191_elite_ultragirl collided on the
number, and the conflicts were the usual "infomaniak and ovh added at the same
spot" across the enum, the config union, the client registry, SENSITIVE_FIELDS,
both providerLabels maps, the icons and the provider form — all resolved by
keeping both.
Regenerated as 0192_light_lake with a current timestamp rather than renaming the
file, so drizzle still applies it on a database that has already run canary's
0191.
restoreRecord had the restore POST and its zone refresh inside one catch. That
was harmless while refreshZone swallowed failures, but the previous commit made
it throw, which brought a new case into that catch: the restore succeeds and
only the publication fails.
The message then told the user the record "has been deleted" and to recreate it
by hand. It exists at OVH, just unpublished, so following that advice duplicates
it as soon as the zone is refreshed.
The two failures are now reported separately. A failed POST still means the
record is really gone and prints what to recreate. A failed refresh after a
successful restore says the record is back but not served yet, and explicitly
says not to recreate it. Either way the original replacement error is kept, so
the user still learns why the type change failed.
Reported by Greptile on #5258.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
refreshZone runs after the record write has already succeeded, so a failure
there means the change exists at the provider but is not being served yet. It
was left unguarded at all six call sites, so the caller saw a bare transport
error and would reasonably read it as "nothing was applied" — while a type
change had already deleted the old record and created its replacement.
Rolling the write back would destroy correct state over a publish failure, and
re-deleting after an uncertain network error is how records get lost. The error
now says what actually happened and how it resolves: the next successful change
to the zone publishes it, or the user can refresh the zone from the OVH manager.
Guarding the helper covers upsertRecord and deleteRecord too, not just the type
change Greptile pointed at.
Reported by Greptile on #5258.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
canary landed its own 0190 (0190_nappy_anita_blake), which collides with this
branch's 0190_even_mastermind. Renaming would not have been enough: drizzle
applies migrations by comparing timestamps against the last applied one, so this
branch's older `when` would have been silently skipped on any deployment that
had already run canary's 0190, leaving the enum without its new value.
Regenerated as 0191_elite_ultragirl with a current timestamp.
canary landed its own 0190 (0190_nappy_anita_blake), which collides with this
branch's 0190_perpetual_red_skull. Renaming would not have been enough: drizzle
applies migrations by comparing timestamps against the last applied one, so this
branch's older `when` would have been silently skipped on any deployment that
had already run canary's 0190, leaving the enum without its new value.
Regenerated as 0191_cool_christian_walker with a current timestamp.
OVH matches access rules per exact path: a `GET /domain/zone/*` rule grants the
subtree but not the bare `GET /domain/zone` that listZones and testConnection
call. Verified against a live account with a consumer key carrying that single
wildcard rule:
GET /domain/zone -> 403 This call has not been granted
GET /domain/zone/ -> 200
GET /domain/zone/{zone}/record -> 200
The form only asked for rights on `/domain/zone/*`, so a token created by
following it could not list zones at all, and the failure surfaced as a bare
"This call has not been granted" that points nowhere.
The hint now lists the five rights verbatim, and a token missing the root one
gets an error that names it instead of echoing OVH's message.
Reported by @narcisonunez on #5258.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
listZones used /1/product, which lego uses but the API docs don't: it is the
legacy singular route, it returns no pagination metadata, and it 401s for at
least some tokens. The documented /1/products returns `total`, `pages`,
`items_per_page` and `page`, and defaults to 15 domains per page, so an account
with more domains than that would silently lose zones.
Switch to the plural endpoint and walk every page. Verified against a live
account: forcing per_page=2 collects all 5 domains across 3 pages with no
duplicates.
Also inline the single-use createdId helper.
Both reported by @narcisonunez on #5257.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Changing a record's type deletes the record then recreates it with the new type,
because OVH's update payload carries no fieldType. If the creation failed the
name was left with nothing and no rollback.
The delete still has to come first, since OVH rejects a CNAME that would sit
alongside other data on the same name. So on a failed creation the original
record is put back from the copy already fetched before the delete, and the
original error is rethrown. If the restore fails too, the error names the record
that has to be recreated by hand.
Reported by Greptile on #5258.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
toSource always writes the apex as ".", but listRecords already accepted "" and
"@" as apex spellings on read. The upsert lookup compared sources strictly, so a
record stored under one of the other spellings would not have matched and the
upsert would have created a duplicate apex record instead of updating it.
Normalize the candidate's source before comparing, so read and match agree.
Reported by Greptile on #5257.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Verified against a live Infomaniak account: the API returns `source: "."` for
apex records, not "" as assumed.
Reading them back produced a doubled dot ("..example.com"), and writing "" meant
an apex upsert never matched the existing record, so it would have created a
duplicate apex record instead of updating it.
toSource now emits "." for the apex and toFqdn accepts ".", "" and "@" so a
hand-written record still round-trips.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Adds OVHcloud alongside Cloudflare, AWS Route53 and Porkbun, following the
existing DnsClient interface in packages/server/src/utils/dns/.
- ovh.ts implements listZones, listRecords, upsertRecord, updateRecord,
deleteRecord and testConnection against the /domain/zone endpoints of the
OVHcloud API, on any of its seven regional endpoints (ovh-eu/ca/us, kimsufi
and soyoustart).
- A new `ovh` value was added to the DnsProviderType enum along with an
ovhDnsConfigSchema (endpoint, applicationKey, applicationSecret, consumerKey)
in the discriminated union, plus the Drizzle migration for the enum change.
- The application secret and the consumer key are masked/merged like the other
providers' secrets in services/dns-provider.ts.
- UI: OVHcloud icon, an endpoint selector and the three credential fields in the
DNS provider dialog, plus registration in the provider selector.
Three OVH-specific behaviours are handled explicitly:
- Requests are signed with `$1$` + sha1(applicationSecret+consumerKey+method+
url+body+timestamp). The timestamp comes from the API's own clock via an
unauthenticated GET /auth/time, since a host clock a few seconds off would get
every call rejected; the measured drift is cached per endpoint for an hour.
- OVH only applies zone changes once the zone is explicitly refreshed, so every
successful create, update and delete is followed by POST /domain/zone/{zone}
/refresh.
- The record update payload carries no fieldType, so changing a record's type
replaces the record (DELETE then POST) and returns the new id.
The record listing endpoint returns ids only, so each record is fetched
individually with the fan-out capped at 8 concurrent requests.
Also fills in the missing Porkbun label in show-dns-providers.tsx, which fell
back to displaying the raw enum value.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Adds Infomaniak alongside Cloudflare, AWS Route53 and Porkbun, following the
existing DnsClient interface in packages/server/src/utils/dns/.
- infomaniak.ts implements listZones, listRecords, upsertRecord, updateRecord,
deleteRecord and testConnection against the Infomaniak API. Zones come from
/1/product?service_name=domain and records from the v2 /2/zones/{zone}/records
endpoints, which are keyed by zone name rather than by product id.
- A new `infomaniak` value was added to the DnsProviderType enum along with an
infomaniakDnsConfigSchema (apiToken) in the discriminated union, plus the
Drizzle migration for the enum change.
- The token is masked/merged like the other providers in services/dns-provider.ts.
- UI: Infomaniak icon and API Token field in the DNS provider dialog, plus
registration in the provider selector.
Infomaniak's `source` is relative to the zone (empty for the apex), so record
names are translated between Dokploy's fully-qualified format and Infomaniak's
subdomain-only format internally, with the trailing dot handled. TXT targets are
stored quoted by the API and are unquoted on read / quoted on write so that
editing a record does not stack quotes on every save.
Also fills in the missing Porkbun label in show-dns-providers.tsx, which fell
back to displaying the raw enum value.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Embedding the full YAML (base64-encoded) into a single SSH exec command
silently fails once the payload passes the SSH transport's packet size
ceiling (~90-100KB), truncating the dynamic config with no error
surfaced anywhere. Switch to SFTP, which has no such limit.
--project-directory was added unconditionally in 87b914996 to fix
relative bind mount resolution for git-based compose deploys with a
nested composePath (#5181). It also moves where build.context and the
generated .env resolve, breaking any compose file in a subdirectory
that has context: . alongside its Dockerfile (#5230), or interpolates
env vars (#5242).
Only pin --project-directory when the compose actually has mounts
configured; otherwise build.context and .env resolve against the
compose file's own directory like plain docker compose. Also pass
--env-file explicitly pointing at the generated .env next to the
compose file, so it's found even when --project-directory is pinned
(credit: tonnenpinguin, PR #5235).
Fixes#5230Fixes#5242
Attaching Basic Auth or a Redirect to an application with no domains
wrote an empty routers/services block to the app's dynamic config
file. Traefik's file provider rejects that as invalid and aborts its
watcher, blocking config updates for every other application.
Fixes#5189
Compose deployments using a git provider (GitHub/GitLab/Gitea/Bitbucket/
custom Git) clone the repo verbatim, so composePath can point into a
subdirectory (e.g. deploy/docker-compose.yml). Docker Compose resolves
relative bind mounts (../files) against the compose file's own directory,
not against the code/ dir where file mounts are actually written, so the
mount silently binds an empty directory instead of the configured content.
Pin --project-directory to code/ when building the docker compose command
so relative mounts always resolve the same way raw deployments already do.
Fixes#5181
Mauricio Siu