mc keeps one --resolve IP per host, and Coolify always pinned the IPv4
address of a dual-stack endpoint. On servers without IPv4, uploads failed
with "network is unreachable". A helper container on the upload network
now connects to each validated IP, IPv4 first, and the first IP that
connects is pinned. S3 restores are now pinned in the same way.
Closes#12397
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- Require write access on the target environment to clone or move a resource
- Check read access on shared variables before a reference is added or resolved
- Limit team credentials, git sources, and imports of other backups to team-wide writers
- Require the delete level for backup, volume, storage, and task deletes
- Require create access for previews, new projects, and git sources
- Use the token team for the access scope, and check team-wide access on Livewire updates
- Require instance admin to toggle the API and MCP server
- Hide server pickers and broadcast details from members without access
Proxies started before v4.4.4 kept the old nginx config after an upgrade:
the upstream IP was resolved only once and IPv6 connections were reset.
- Label new proxies with a configuration version
- Status checks (Sentinel and SSH) recreate a running proxy with an
older or missing version once, at most every 10 minutes
- Remove the old proxy only after the image pull, so a failed pull
keeps it running
Closes#10198
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Add a "Use path-style URLs" setting to S3 storages. It is detected from
the endpoint (Alibaba OSS, Tencent COS, Volcengine/BytePlus TOS), and the
connection test switches to virtual-hosted style when the provider
rejects path-style requests. Backups pass `--path off` to mc.
Closes#10347
Resolve conflicts between main's security fixes and v4.5 features:
- Keep v4.5 features (log redaction, multi-S3 backups, transfer flow,
scheduled task modal, sidebar) and port main's fixes into them.
- Reuse RequiresDeployAbility for the Hostinger create endpoint.
- Align MCP tool abilities and descriptions with main's new deploy
requirements, and use deploy tokens in the affected MCP tests.
- Drop the duplicate DockerCleanupJobTest helper and tests, and the
exportBundle test for a method v4.5 removed.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A custom config file written with tee gets the server umask. With umask
027 or 077 the file is not readable by the container user, so MySQL and
MariaDB ignored the whole config without an error.
- Set mode 644 on the MySQL, MariaDB, PostgreSQL, and MongoDB custom
config files after each write.
- Close the multiplexed SSH connection when the server user, IP, or port
changes, so the next command does not run as the old user.
Closes#11600
- Stop the Redis username accessor from creating REDIS_USERNAME on read
- Keep source values when cloning applications (no duplicate NIXPACKS_NODE_VERSION)
- Match exact keys when adding default database environment variables
- Write content files before database containers start
- Resolve the database proxy upstream through Docker DNS on each connection
- Remove existing duplicate REDIS_USERNAME and NIXPACKS_NODE_VERSION rows
Docker publishes the proxy port on [::] too and forwards it to the
container's IPv6 address when the network has IPv6 enabled. nginx
listened only on IPv4, so every IPv6 connection was reset.
Add listen [::]:<port> when the destination network has IPv6 enabled.
Networks without IPv6 keep the IPv4-only listener.
Fixes#12133
v4.3.23 started Redis with the stored REDIS_PASSWORD value as it is and
unquoted, and the connection URLs used the same value. 4.4.0 resolved a
shared variable reference for the start command only, so after the first
restart the server password changed and no longer matched the URLs.
- Add legacy_password_quoting to standalone_redis; existing databases
keep the v4.3.23 server password
- New databases resolve REDIS_PASSWORD (local, shared or remote secret)
and quote it in the start command
- Connection URLs use the same password as the server
- Hide the flag in the API, as for KeyDB and Dragonfly
The stored password is the default. Before, a local REDIS_PASSWORD was
rewritten to the stored password, and a shared one replaced the stored
password, so the old password was lost.
- A local, shared, or remote secret REDIS_PASSWORD sets the server
password; the stored password and the variable are not changed
- Removing the variable, or an empty value, uses the stored password
- Connection URLs show the password in use; for a remote secret they
show the reference, not the secret
- Databases created before this release keep the v4.3.23 behavior
Scope resource permissions and related records to the owning team across
team switches, cloning, deployments, and resource operations.
Merge ClickHouse volume mounts, preserve anonymous data volumes on stop,
and require adopting current data before starting or restarting.
Fix file storage deletion checks and add regression coverage.
Require deploy or root token access when creating managed mounts outside
resource directories or changing their file contents through the API.
Synchronize KeyDB and Dragonfly passwords with local and shared
REDIS_PASSWORD values while preserving secret manager resolution.
Allow validated host file and directory mounts outside resource directories
while protecting external paths from deletion.
Clean up cancelled deployment containers and clear tracked process IDs.
Preserve legacy KeyDB and Dragonfly password parsing, accept server backup
paths without extension restrictions, and allow Compose variable proxy ports.
Addresses CodeRabbit review on #11513:
- Add Livewire feature tests for Heading::flush() covering success +
audit event, non-cache guard, denied manage permission, and menu
visibility for cache vs non-cache databases.
- Document the pre-existing parseCommandsByLineForSudo() limitation that
mutates passwords containing shell sequences on non-root servers.
Adds a Flush cache action to the database heading for the three
Redis-protocol cache databases. It runs FLUSHALL ASYNC inside the
running container to clear all keys without stopping the container.
The option is hidden for SQL/Mongo databases and only shown while the
database is running. A confirmation modal requires typing the database
name before flushing. The redis-cli password is shell-escaped to
prevent command injection.
A stop or delete no longer queues a full cleanup each time. It skips the
cleanup when one finished on the server in the last hour, and the queued
run only cleans when disk usage is at or above the server cleanup
threshold. The docker_cleanup API flag keeps its default of true.
Cloud production now runs one supervisor per queue (deployments, crons,
high, default, maintenance), so a busy queue cannot starve the others.
Add maintenance_queue() for slow Docker cleanups and a shared
horizon.worker_timeout config value. Self-hosted keeps the s6 supervisor.
On cloud, StartService and DeployServiceApplication dispatched their
CoolifyTask on the high queue, so service starts competed with all other
high jobs. remote_process() and CoolifyTask now accept a queue (default
high); start paths pass deployment_queue(). The queued RestartDatabase
action also runs on deployment_queue() instead of the default queue.
The start command now quotes the password. Databases created before this
release keep the unquoted v4.3.23 argument when quoting would change the
password the server receives, until their password changes. Remote secret
passwords are always quoted exactly.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- Write remote secret values as exact dotenv literals in the service .env
file and escape "$" where they go directly into a generated compose file
(environment, start commands, healthchecks).
- Resolve remote secrets before a service or database restart stops the
running resource, and show the secret manager error instead of a generic
start failure (new RemoteSecretException).
- Cache fetched key names for the autocomplete and rate limit how often a
user can contact the secret manager.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
An import task can now run until the SSH command timeout (capped below the
Horizon worker timeout). When the CoolifyTask fails, times out, or its SSH
command ends with 124/255, Coolify stops the restore in the database
container and keeps the import in progress until the stop has run, so the
database stays blocked. The restore reports its own 124/255 as 1.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The MongoDB health check passed only on the writable primary
(db.hello().isWritablePrimary or isMaster().ismaster). A replica set
that was never initiated, which is common for Prisma setups, and any
non-primary member were reported unhealthy after a restart.
The check now runs db.adminCommand({ ping: 1 }) with mongosh or the
legacy mongo shell. Ping needs no authentication, passes on every
running member, and still fails when mongod is not reachable.
The health check from #11730 picked mongosh or the legacy mongo shell
from the image name (mongo:4*). Images without mongosh whose name does
not start with mongo:4, such as mongo:3.6 or registry mirrors of
mongo:4.4, ran a missing mongosh (exit 127) and stayed unhealthy.
The check now uses mongosh when the container has it and falls back
to the legacy mongo shell otherwise.
Replace the test that only checked the source text with tests that run
the generated script with fake shells.
- Regenerate SSL Certificates now writes server.crt/server.key for every
engine except MongoDB, which keeps the combined server.pem. Before, the
button always wrote server.pem, so Postgres, MySQL, MariaDB, Redis, KeyDB
and Dragonfly failed to start after a regenerate.
- The certificate renewal job uses the same rule, so MongoDB no longer
switches to server.crt/server.key on renewal and fails on restart.
- Starting an SSL database writes the CA that signs its certificate to
/data/coolify/ssl/coolify-ca.crt, so a stale CA file (for example one
written by another server entry for the same host) no longer causes
"certificate signature failure".
- Saving a custom CA certificate that does not match the stored CA
private key is rejected.
- The CA file write commands live in SslHelper::caCertificateFileCommands().
Fixes#10388Fixes#6771Fixes#11926
An all-databases MySQL/MariaDB restore also restored the mysql system
database, so after the next restart the root and app passwords were the
source ones and the MariaDB health check failed. The restore now skips mysql,
sys, performance_schema and information_schema by default. The new option
"Restore users and privileges" (restore_mysql_users in the API) keeps the
old behaviour.
A queued start or restart whose reservation expired no longer runs. Failed
queued starts are recorded as failed activities and release the reservation.
- database import: add keep_owners (restore PostgreSQL owners and
privileges) and sqlite_database (pick the target SQLite file) to the
API, OpenAPI spec, and import form
- oauth: require password confirmation and support team selection on
OAuth login
- remote process: send team broadcast events to the team of the server
that ran the process instead of the causer's user id
- servers API: handle the legacy build server setting
- compose: support parent bind paths and volume driver options
- dns: improve record hints and provider management
- notifications: fix throttling and missing backup notifications
- ui: rework the confirmation modal
- add feature and unit tests for the new behavior
A non-root SSH user cannot open redirects, expand globs, or cd into
/data/coolify (9999:root 0700). Several remote commands relied on that
and failed on non-root servers.
- Write README files with `tee` instead of shell redirects in database
starts and application deployments
- Clear Postgres init scripts with `find` instead of a glob
- Split the proxy config backfill if/else so each branch gets sudo
- Start the log drain with `docker compose -f` instead of `cd`
- Wrap the preserved-repository compose start in `sh -c` for non-root
- Stage clone archives outside /data/coolify via StagesCloneArchives
- Extend the sudo parser and add non-root command tests
Move the shared team id and team channel logic of 15 broadcast events
into a BroadcastsToTeam trait. Events no longer fall back to the session
team: the session team can differ from the resource team, and queued
jobs have no session.
Fix StopDatabaseProxy, which dispatched DatabaseProxyStopped without a
team, so the event was never broadcast from queued jobs.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
peaklabs-dev
Andras BacsaiandClaude Opus 5.5
