When another process held the control socket, `ssh -fN` disabled
multiplexing and stayed open as a plain connection forever. Each one
kept an sshd process on the server, which used up its memory.
Run `true` instead of -fN. ControlPersist moves a new master to the
background; a process that cannot become the master exits.
Public pull request previews no longer get production values, and no
preview falls back to production variables. Existing previews were
deployed with these values; for example, their databases were
initialized with the production password, so they broke on redeploy.
Previews that exist at the upgrade get keeps_production_values and
keep the earlier behavior. Previews created later do not. Server
transfer moves the flag.
The retry labels (Traefik retry middleware, Caddy lb_try_duration)
changed the labels of every application. On the first rolling update
after a settings save, the old and the new container defined the same
Traefik router differently, so Traefik dropped the route and returned
503 until the old container stopped.
Remove the retry labels from this release. The rolling update drain
wait stays.
Webhook secrets are stored settings, so a token with the write ability can
set or change them, the same as GitHub App and GitLab App webhook secrets.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
- Add an Axiom edge domain setting (default us-east-1) and use the /v1/ingest endpoint, so EU datasets work
- Remove the Highlight drain (service ended) and stop its running containers
- Update Fluent Bit from 2.0 to 5.1 and use the warn log level
- Restart running drains when the generated files change, with the new image pulled before the old container stops
- Show log drain validation errors under their fields
Closes#9898
Cloning a clone added another "-clone-<id>" to the name, so after about
seven clones the name was longer than 255 characters and the insert failed.
Closes#9976
The log endpoints found containers with `docker ps -a`, then rejected
every container that was not running. Remove that check, so failed
deployments can be diagnosed through the API. Without service_name,
the application endpoint prefers a running container. The old error
stays when no container exists.
Closes#9972
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The analytics page fetched each server's dashboard bundle over SSH one
after another, so the skeleton stayed until every server answered. Start
all fetches at the same time and remember failures, so the page waits
only for the slowest server.
Closes#12347
Provider errors now return 502 with a safe message, the provider status, and
the provider error code instead of a bare 500. Hetzner no longer sends a create
request again after a 5xx, and a VPS that Coolify did not save is deleted.
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
Andras BacsaiandClaude Opus 5.5