Replaces the `DOKKU_ARCHIVE_MAX_SIZE` and `DOKKU_ARCHIVE_MAX_FILES` environment variables with global git properties (`archive-max-size` and `archive-max-files`), configurable via `dokku git:set --global` and surfaced through `dokku git:report --global`. Defaults remain `1073741824` bytes and `10000` entries.
Archives passed to git:from-archive and certs:add were extracted without symlink or path validation, allowing a crafted archive to write arbitrary files anywhere writable by the dokku user via symlink traversal. Extraction now pre-scans entries for absolute paths, parent traversal, and unsafe symlinks, applies the GNU tar `--no-unsafe-links` flag when available, and validates symlinks after extraction.
The previous use of `touch` before `netrc set` allowed the file to inherit the umask and be world-readable, exposing stored git credentials to local users. The set and unset paths now explicitly chmod 0600 and chown to the dokku user, and the plugin install hook repairs permissions on already-affected installations.
Add defense-in-depth sanitization for OpenResty include files to prevent
OS command injection via malicious filenames that break shell quoting in eval.
- Add filename validation in core-post-extract using regex [^a-zA-Z0-9_.-]
- Validate both http-includes and location-includes paths
- Abort deploy via dokku_log_fail on unsafe filenames
- Skip non-regular files (symlinks, directories) during extraction
- Add security regression test with unsafe filename containing space
- Keep existing guards in docker-args-process-deploy as belt-and-suspenders
- Update documentation to clarify allowed filename characters
Addresses CVSS 9.9 vulnerability where filenames like poc'$(cmd)'x.conf
could escape shell quoting and execute arbitrary commands during deploy.
The builder-dockerfile, builder-lambda, builder-nixpacks, builder-pack and builder-railpack `core-post-extract` triggers assigned `$2` to a local `SOURCECODE_WORK_DIR` but called `pushd "$TMP_WORK_DIR"`, which was unset. Bash 5.2 silently accepted `pushd ""`, so the bug stayed dormant. Bash 5.3 (shipped with Ubuntu 26.04) makes it a hard error and `set -e` aborts the trigger, causing every `git push` to fail with `pushd: null directory`.
The release_and_deploy function and several proxy plugin scripts call fn-plugin-property-* helpers without sourcing `plugins/common/property-functions`, producing `command not found` errors during deploys. Source the helpers from `plugins/common/functions` and from each proxy plugin entry point that consumes them.
`SaveEntry` and `touchMigrationFlag` wrote files via `os.WriteFile` and `os.Create` and never chowned them, so the install-time legacy-mount migration produced root-owned `legacy-*.json` files in `/var/lib/dokku/data/storage-registry/entries/`. The dokku user that runs `ps:rebuild` could not read them and every rebuild on a 0.37.x to 0.38.0 upgrade failed with permission denied. The companion `repairRegistryOwnership` pass in `TriggerInstall` rewrites ownership across the whole registry tree so installs that already ran the buggy code are fixed on the next package upgrade, since the per-app migration flag would otherwise cause `MigrateLegacyMounts` to skip the broken files forever.
The argument-handling block in each plugin's report command overwrote APP with --global before checking whether $1 actually held an info flag, so `dokku <plugin>:report --global --<plugin>-foo` always emitted the full report instead of the requested value. Reorder the conditionals across the sixteen plugins that share the pattern so the info-flag check runs first, and add a regression bats case for each one.
The byjg/easy-haproxy image polls Docker for label changes every 10
seconds by default, which races with the haproxy bats suite and
intermittently produces curl exit 7. Expose `refresh-conf` as a
global-only haproxy property that maps to `EASYHAPROXY_REFRESH_CONF`,
lower it to 2 seconds in the bats setup, and wrap the localhost HTTP
assertions in a retry loop so checks wait for haproxy to converge
rather than failing on the first attempt.
The default nginx template rendered four near-identical server blocks per app, so the same listen, access_log, error_log, ssl_*, error_page, and proxy chain had to be maintained across http, https, grpc, and grpcs branches. Merging http and https into a single branch keyed on an `is_ssl` boolean, and likewise grpc and grpcs, removes the duplicate proxy_set_header chain and error-page locations and brings the structure in line with the openresty proxy template. Output is preserved up to whitespace and the existing `cat -s` pass already squashes the leftover blank lines. Adds plugins/nginx-vhosts/template_test.go exercising the rendered output via sigil as a Go library across HTTP-only, HTTPS, HTTP-to-HTTPS redirect, no-listeners 502 fallback, gRPC, gRPCs, gRPC-without-listeners skip, IPv4 bind, upstream blocks with and without keepalive, the access_log `off` short-circuit, the X-Forwarded-Ssl toggle, the http2 listen-parameter vs directive split, http2_push_preload conditional emission, and the nginx.conf.d/*.conf include in every code path; and tests/unit/nginx-vhosts_16.bats covering deploy plus `nginx -t` end-to-end for both HTTP and HTTPS apps.
# History
## 0.38.0
Install/update via the bootstrap script:
```shell
wget -NP . https://dokku.com/install/v0.38.0/bootstrap.sh
sudo DOKKU_TAG=v0.38.0 bash bootstrap.sh
```
See the [0.38.0 migration guide](/docs/appendices/0.38.0-migration-guide.md) for more information on migrating to 0.38.0.
### Bug Fixes
- #8533: @josegonzalez Split env config and image pull secret into separate helm releases
- #8530: @josegonzalez Split multi-flag input in docker-options
- #8528: @josegonzalez Skip retiring images still in use by app containers
- #8525: @josegonzalez Add launcher entrypoint for CNB images on dokku run and cron:run
- #8522: @josegonzalez Only emit keda fallback when a non-cpu/memory trigger exists
- #8515: @josegonzalez Fix vector mount directory config
- #8508: @josegonzalez Preserve all domains when renaming an app
- #8507: @josegonzalez Retire orphaned containers when scaling down
### New Features
- #8538: @josegonzalez Add scheduler-aware named storage entries
- #8527: @josegonzalez Accept --global on :report subcommands
- #8524: @josegonzalez Pre-validate custom nginx.conf.sigil during core-post-extract
- #8523: @josegonzalez Support resource limits on the build container
- #8517: @josegonzalez Send SIGTERM to old containers immediately on deploy
- #8516: @josegonzalez Scope docker-options to specific procfile processes
- #8509: @josegonzalez Ship default catch-all site on fresh apt install
- #8506: @josegonzalez Add --format json to git:report and nginx:report
- #8505: @josegonzalez Add git:auth-status to check netrc match
- #8493: @josegonzalez Generate 502 config for apps without web listeners
- #8404: @josegonzalez Upgrade vector chart from 0.42.0 to 0.52.0
- #8403: @josegonzalez Upgrade ingress-nginx chart from 4.10.0 to 4.15.1
- #8402: @josegonzalez Upgrade keda to 2.19.0 and keda-add-ons-http to 0.12.2
- #8259: @josegonzalez Add post-create support for env key in app.json
- #8157: @josegonzalez Add support for specifying buildpacks via app.json
- #8154: @josegonzalez Enable live-restore by default when installing Dokku
- #3697: @josegonzalez Migrate builds plugin to go and track per-build records
### Refactors
- #8514: @josegonzalez Migrate docker-options subcommands to go
- #6716: @josegonzalez Move app and global ENV files to consolidated config path
### Dependencies
- #8541: @dependabot[bot] chore(deps): bump traefik from v3.6.14 to v3.6.15 in /plugins/traefik-vhosts
- #8537: @dependabot[bot] chore(deps): bump github.com/traefik/traefik/v2 from 2.11.43 to 2.11.44 in /plugins/scheduler-k3s
- #8535: @dependabot[bot] chore(deps): bump github.com/onsi/gomega from 1.39.1 to 1.40.0 in /plugins/common
- #8529: @josegonzalez chore: bump dokku/netrc to v0.11.0
- #8520: @dependabot[bot] chore(deps): bump packaging from 26.1 to 26.2 in /docs/_build
- #8510: @dependabot[bot] chore(deps): bump packaging from 26.1 to 26.2 in /docs/_build
- #8503: @josegonzalez Bump dependency versions and add daily updater workflow
- #8502: @josegonzalez Bump go version to 1.26.2
- #8495: @dependabot[bot] chore(deps): bump k8s.io/apimachinery from 0.35.4 to 0.36.0 in /plugins/scheduler-k3s
- #8494: @dependabot[bot] chore(deps): bump dokku/openresty-docker-proxy from 0.9.3 to 0.10.0 in /plugins/openresty-vhosts
- #8490: @dependabot[bot] chore(deps): bump k8s.io/kubernetes from 1.35.4 to 1.36.0 in /plugins/scheduler-k3s
### Other
- #8498: @josegonzalez Migrate environment variables to plugin properties
Adds plugins/storage/migrate_test.go exercising migrateApp and MigrateLegacyMounts: single deploy phase, cross-phase grouping into one attachment, :ro and free-form volume options preserved on the attachment, idempotency on a second pass, the per-app flag-file fast path that skips already-migrated apps, and the name-collision refusal that aborts before draining. tests/unit/storage.bats gains an end-to-end case that stages a -v line via docker-options:add, runs storage:migrate, and asserts the synthesized colon form shows up in storage:list, the legacy-<hash> entry shows up in storage:list-entries, and the original -v line is gone from docker-options:report on both phases. To make that test (and operators with restored backups) practical, storage gains a storage:migrate <app>|--all subcommand backed by a new MigrateApp helper that skips the per-app flag file so a re-run actually re-scans.
Adds typed JSON build records under data/builds/<app>/<build-id>.{json,log} keyed on a stable base36 ULID-style DOKKU_BUILD_ID generated for every deploy. The new commands surface that history (builds:list, builds:info, builds:prune) and an operator-configurable retention via builds:set retention. The existing builds:cancel and builds:output now key on the build-id (with safe handling for already-finalized and abandoned records), and the per-build log file replaces journalctl as the durable source of truth for builds:output.
CallExecCommand wraps non-zero exit codes as errors with the code populated on the response, so storage:exec was returning the wrapped error and being collapsed to exit 1 by LogFailWithError before the os.Exit branch ran. Both storage's CommandExec and scheduler-docker-local's TriggerSchedulerStorageExec now check ExitCode before err so the underlying tool's status flows through verbatim. Separately, the legacy host:container colon form of storage:mount used to write straight into docker-options, but storage:list now reads only attachments, so newly mounted legacy-form storage was invisible. CommandMount and CommandUnmount route the colon form through LegacyMountToEntry plus AddAttachment / RemoveAttachment, making storage:list show every mount regardless of form while preserving the existing "Mount path already exists." / "Mount path does not exist." error wording.
The install trigger created /var/lib/dokku/data/storage-registry and its entries / migrations subdirectories as root with mode 0755, so the dokku user that runs storage:create couldn't write entry JSON files or migration flag files into them. Mirrors what PropertySetup does for config/storage by calling SetPermissions on each registry path. This unblocks the bats coverage that was failing at the first storage:create call.
Moves the actual exec out of the storage plugin and into a new scheduler-storage-exec plugn trigger. scheduler-docker-local does docker run with TTY-aware -it/-i selection and --user derived from entry.Chown; scheduler-k3s creates a throwaway Pod via the kubernetes API, waits for it to reach Running with structured error reporting (ImagePullBackOff and friends are surfaced from the container status verbatim, no kubectl involvement), execs the user command via the existing SPDY plumbing in k8s.go, and deletes the Pod on the way out. (Entry).Validate now accepts either an absolute path or a docker named-volume token for docker-local entries so the migration synthesizer's named-volume legacy entries work cleanly. storage:exec gains --as-user for one-off uid overrides, propagates the underlying tool's exit code via os.Exit, and detects TTY/interactive mode from os.Stdin so non-interactive scripted use no longer trips over docker's input-device-is-not-a-tty error.
storage:list was calling the storage-list plugn trigger which read -v lines from docker-options. After the install-time migration drains those lines into the attachment store, that source is empty for every migrated app and for any app that only ever used storage:create + storage:mount. The fix moves CommandList to call a new in-process ListAppMountEntries helper that reads attachments directly, surfaces the entry name in JSON output via a new entry_name field, and falls back to the entry name as the host token for k3s entries with no host path so the colon form remains well-formed. The storage-list plugn trigger is kept for back-compat with external callers but emits a deprecation warning and now reads from the same attachment-driven source.
Switches the storage plugin's post-delete and post-app-rename-setup triggers from rewriting the attachment list as empty to removing the per-app property folder outright via PropertyDestroy. This matches the convention other plugins use on app deletion and avoids leaving an empty mounts file behind for an app that no longer exists.
apps:destroy was failing in CI because the post-delete trigger called PropertyListWrite, which tried to mkdir /var/lib/dokku/config/storage/<app> as the dokku user but the parent directory had been created by the install trigger as root. The install trigger now calls PropertySetup("storage") so the per-plugin config root has the correct ownership, and post-delete / post-app-clone / post-app-rename short-circuit when the app has no attachments to begin with so they can't fail on a never-touched property tree. The entry registry and migration flag directories also moved out of config/storage (where they would have collided with the property-store path config/storage/<appName>) into a dedicated data/storage-registry tree.
storage:set is the in-place editor for an existing entry; size grows are passed through to a helm upgrade and Kubernetes-incompatible swaps (access-mode, storage-class) are rejected up front. storage:exec spawns a throwaway container that mounts the entry: docker run on docker-local entries, kubectl run with a Pod-spec override on k3s, falling back to bash then sh when no command is supplied. storage:wait blocks until a k3s entry's PVC reports Bound via the storage-status trigger, with a docker-local fast path that just checks the host directory. storage:report --global enumerates every registered entry and the apps that mount it, in text or JSON form.
Adds a per-entry storage helm chart that owns the PVC and (when host_path is set) the PV, and three triggers for the storage plugin to drive: storage-create installs/upgrades the chart with storage-class existence validation, storage-destroy uninstalls it, storage-status reports the PVC phase. App deployment and cron-job charts now consume storage-app-mounts and reference each PVC by name; the // todo: implement volumes blocks are gone. ProcessVolume gained sub_path, read_only, and persistent_claim fields and the deployment/cron-job templates render them.
storage:mount and storage:unmount now accept the named-entry form (--container-dir, --phase, --process-type, --volume-subpath, --volume-readonly, --volume-chown), persisting attachments via the property list. The legacy host:container colon form is preserved for back-compat. Storage now emits its own docker-args-deploy and docker-args-run triggers so docker-local apps continue to receive bind mounts even after their data moves out of docker-options. The install trigger walks every app once per upgrade and converts legacy -v lines into legacy-<hash> entries plus attachments, draining the originals from docker-options last so a partial run leaves the old behavior intact. Lifecycle triggers post-delete, post-app-clone-setup, and post-app-rename-setup keep the attachment store in sync with app state.
Wires the new entry-management commands on top of the Entry type. storage:create accepts both docker-local entries (with idempotent host-directory creation) and k3s entries (which delegate PVC provisioning to a forthcoming scheduler-side storage-create trigger). storage:destroy refuses to remove an entry that any app still mounts. storage:ensure-directory now emits a deprecation warning. storage:mount and storage:unmount keep the existing colon-form behavior so this commit is non-breaking; the named-entry mount form follows in a later commit.
Introduces the Entry and Attachment types that the storage plugin will use as the source of truth for named storage volumes, replacing the colon-delimited mount strings stored under docker-options. Entries persist as JSON under config/storage/entries and validate against scheduler-specific rules; attachments persist via the property list and capture how an app uses an entry. The legacy migration name synthesizer is also added so existing colon-form mounts can converge on a deterministic legacy- entry name. Existing storage commands are unchanged in this commit.
This still needs tightening to ensure we don't trigger this for normal git plugin commands - as well as anything that triggers a deploy - but works in a pinch.
The DOKKU_PID now never gets overwritten except in the case that DOKKU is executed by the sudo user. If the command ends up executing a deploy, then the pid of the `dokku` owned process - which may have been executed via sudo - will be written to the file lock, allowing future commands to interact with the original process.
Additionally, the new builds plugin can be used to handle killing a build.
Multi-flag inputs (e.g. `--build-arg X=Y --link a --link b`) used to be stored as a single line, which bypassed the per-line filter that drops `--link` and similar flags for dockerfile-based builders. Each `--flag [value]` group is now stored as its own entry, and a `--process` typed after the app name is lifted into the subcommand flag instead of being stored as a docker option.
The new config and pull secret helm releases need to honor user-set annotations and labels for `--resource-type secret` so existing scheduler-k3s annotation tests keep passing. The dedicated charts now render `.Values.global.annotations` and `.Values.global.labels` onto their Secret manifest, and the deploy trigger plumbs `SecretAnnotations` and `SecretLabels` from the global annotation/label config. The rollback regression bats test now uses `dokku ps:rebuild` for its second deploy because git push of an unchanged ref is rejected by the dokku remote.
