Require deploy permission for API changes that trigger server side effects, including resource deletion, storage access, provider operations, and database imports. Also escape Docker cleanup filters and sanitize log drain server names.
oldRawParser made the bind directory of a long-syntax source like
../data at {workdir}./data, next to the path that Docker Compose uses.
Resolve the source with replaceLocalSource(), as the compose parsers do,
so Coolify makes the directory that the container mounts. A source that
resolves above / now stops the deployment before remote work.
Closes#4055
Docker Engine 27.0 to 28.0.0 reports an IPv6 network gateway with its
prefix length (fd00::1/64) until the daemon restarts (moby/moby#49520).
The Docker CLI 29 and Compose 5 in the helper image reject it, so
deployments failed and the upgrade left Coolify stopped.
- Stop the upgrade before it changes containers when the coolify network
has such a gateway, and tell the user to restart or upgrade Docker.
- Create a new coolify network again without IPv6 when Docker reports
its gateway with a prefix.
- Report an error, not "Upgrade complete", when the new Coolify
containers do not start.
- Add the fix to deployment errors that contain this ParseAddr failure.
Fixes#11583
Merge log rotation and the network pool into the existing daemon.json
instead of overwriting it. Skip the change and the Docker restart when
containers are running, unless DOCKER_POOL_FORCE_OVERRIDE=true.
Closes#11221
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Skip the pre-stop pull when a custom build command runs without a build server, and pull only services whose image does not come from a build, including services that reuse the image of a build service.
Closes#11773
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
install.sh created the resource directories and ran chown -R and
chmod -R on source, ssh, and images before step 1. When a check or a
CDN download failed, the changed owners and modes stayed although
nothing was installed.
Create only /data/coolify and source (for the log and the downloads)
at the start. Create the other directories and set the permissions in
the SSH key step, after the checks and downloads succeed.
Closes#11973
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
validateOS matched only ID from /etc/os-release, so derivatives such as
Omarchy, CachyOS, Manjaro or Linux Mint failed validation before the
Docker checks ran. Check ID first, then each ID_LIKE entry, and keep the
exactly-one-match rule. Move the parsing into a testable function.
Closes#12033
Co-authored-by: Furox <177975472+Furox-Art@users.noreply.github.com>
The repository URL rule rejected ssh://user@host:port/path, so deploy-key
resources on servers with a custom SSH port needed a manual URL edit after
creation. Accept ssh:// URLs and keep them unchanged for deploy keys so the
port reaches the clone. Rename scpStyleGitUrlToHttps to sshGitUrlToHttps and
let it convert ssh:// URLs too, so public flows, repository links and GitHub
App repository slugs handle them.
Closes#12161
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Count invalid keys, tokens, and deleted webhooks in a new failure_count column
on each notification settings table, like servers.unreachable_count. After 3
failures in a row the channel is disabled and the team is told by instance
email and its other channels. Rate limits and server errors are retried and
not counted; test notifications never count.
Applications and service applications can now use the server resolver,
no resolver ("none"), or a custom resolver name. This allows servers
that mix Let's Encrypt domains with domains that use their own
certificates, such as Cloudflare Origin CA. The setting is available in
the UI and the API.
Closes#12138
Use the helper's RAILPACK_VERSION, falling back to the installed CLI
version, instead of a hardcoded frontend version. Add regression tests
for both version sources and application-variable isolation.
Store server IPs in one form (no brackets, short lower-case IPv6) and
write each address in the form its context needs:
- URLs and host:port (Sentinel, database URLs, SFTP backup download,
DNS servers, SMTP, webhook base URLs, links, health checks) get
brackets through formatHostForUrl().
- Traefik Host() rules use the address without brackets, so IPv6
domains no longer return 404; saved labels are fixed at deploy.
Caddy keeps the brackets.
- SSH keeps the plain address.
- Hetzner gives an IPv6 /64 network; IPv6-only servers now get the
server address (prefix::1) instead of the unusable prefix::64.
- sslip labels, registry image names with an IPv6 host, the API domain
list, allowlist matching and duplicate checks handle IPv6.
A migration normalizes stored IPv6 server addresses, repairs Hetzner
servers saved as ::64 that are not reachable, and adds brackets to
saved Sentinel URLs.
An upgrade pulls new images and recreates the Coolify containers. A full
disk during an upgrade can leave Coolify broken.
- Measure free space on the Docker root dir and /data/coolify of localhost;
require 5 GB. An unknown value does not block.
- Manual upgrade: show the free and required space, with Run Docker cleanup,
Check again, and a confirmed Upgrade anyway override.
- Automatic upgrade: skip it and notify the team (disk usage channels, at
most once per 24 hours per server).
- upgrade.sh: stop before any change when the disk is low. A new 5th
argument skips the check (used by the override).
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Accept HTTP(S), git:// and SCP-style SSH repository URLs whose host is an
IP address, including bracketed IPv6 such as git@[2a01:4f8::1]:org/repo.git.
The clone runs on the user's deployment server, so the IP block did not
protect Coolify, and SSH URLs with IPv4 already passed.
Loopback, link-local and other reserved addresses are still rejected as
internal hosts. Brackets are allowed only around a valid IPv6 host.
Accept the `?`, `:?`, `+` and `:+` interpolation operators in Compose
network names, such as `${CONTAINER_NAME?error}`. The validation added
in v4.4.0 only knew `-` and `:-`, so these valid files stopped deploying.
The text after `?` is an error message and is not used as the default of
the created environment variable. Unsafe operands are still rejected.
Fixes#12126
Block new upgrades while a recent upgrade is running, allowing retries after 15 minutes without a status update.
Add request timeouts and polling guards to keep upgrade progress responsive during restarts.
Pull request deployments sent every non-railpack build pack to build_image(),
which no longer handles the static build pack since it was moved to
build_static_image(). Static preview deployments then failed with
"open Dockerfile: no such file or directory".
Closes#8238
Replace datalist and searchable-listbox components with the shared listbox and migrate form fields.
Deduplicate superseded Livewire updates to prevent repeated array update hooks, with regression coverage.
Sentry self-hosted (getsentry/self-hosted 26.8.0) could not be deployed
as a Docker Compose application. Three separate problems blocked it:
- YAML collection aliases: the first fix raised the limit to 256 only in
Application::oldRawParser(). The safety check added in v4.4.0
(validateDockerComposeForInjection) and all other Compose parse calls
still used the Symfony default of 128, so the Sentry file (133 aliases)
failed with "Maximum number of collection aliases (128) exceeded".
Add parseDockerComposeYaml() and use it for every Compose parse.
- Argument list too long: the deployment job sent the generated Compose
file as base64 inside the SSH command. Linux limits one argument to
128 KiB, and the Sentry file is about 370 KB. execute_remote_command()
now accepts an 'input' key that streams content over SSH stdin, and
the job writes the Compose file with it.
- A list-style build arg without a value (build.args: [SENTRY_IMAGE])
merged into a map environment kept its numeric key, so Docker Compose
rejected the file with "non-string key in ...environment: 0". It now
becomes NAME with no value, as in Compose.
Closes#11725
Count running scheduled tasks, database backups, and volume backups,
excluding runs older than their job timeout. Report scheduled job counts
and keep the progress check active while any scheduled jobs are running.
- Update tests whose expectations no longer matched intended changes
(GitHub App redirect, wildcard domains, container IDs, nightly
install folders, localhost metadata, Sentinel restart on save)
- Seed the instance settings the 2FA profile test needs
- Run the global Pest hooks for all test folders and release the
Laravel application after each test, so tests do not leak state
- Replace a class alias mock that broke later tests
- Generate factory names that always pass name validation; Faker names
with apostrophes or short words made form tests fail at random
Since aabd12737 a Compose bind volume with a content block had to stay
inside the resource directory. Services and Compose applications saved
in v4.3.23 with other paths, and Git Compose files with them, no longer
started or deployed.
Only admins and owners can edit Compose files, and they can already
write any host file with a plain bind mount, so the rule added no
security.
- Write content files to their host path again, as in v4.3.23
- Keep shell escaping and the safe-character checks on volume sources,
so injection attempts are still rejected
The sudo parser read everything after `mkdir -p` as the path, so the
ownership step ended up after the last command of the line. For
`mkdir -p /data/coolify/proxy/dynamic && ls ... | head -n 1000`, `head`
got the `find` options and the proxy dynamic configurations failed on
every server with a non-root user.
- Add the ownership step right after the mkdir path; commands after
`&&` or `;` stay unchanged
- Lines with other forms (redirects, `||`, pipes) get no ownership step
Jobs that v4.3.23 put in the queue have no type property. The promoted
constructor property had no default, so these jobs failed with "must not
be accessed before initialization" after the upgrade.
- Declare type as a normal property with a null default
- Hide proxy configuration contents from users without update permission.
- Select and validate backup S3 storage against the resource's team.
- Keep scheduled branch syncs green while a conflict PR remains open.
- Document external mount permissions and Sentinel traffic limits.
- Add regression coverage for authorization and backup storage selection.
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.
Andras Bacsai
🏔️ Peak