0.4.6
Featuresâ
- Added support for selecting Debian distributions in Debian repositories.
- The bundled RepoFlow documentation now dynamically loads the server URL to improve the user experience when working with the RepoFlow API.
- Added new
GRYPE_AUTO_UPDATE_CRON
environment variable to control how often Grype's vulnerability database is updated. - New env -
ALL_REQUESTS_TIME_OUT_IN_SECONDS
- Docker labels defined inside the Dockerfile now override those inherited from base images or added via build commands, allowing better control over image metadata.
Bug Fixesâ
- Fixed an issue where uploading certain Docker images failed when using the API or UI. Regular
docker push
was not affected. - Fixed an issue where vulnerability scanning failed on some Docker images.
- Fixed an issue where selecting a distribution in the UI during Debian package upload caused a bad request. The API was not affected.
- Introduced a new environment variable:
ALL_REQUESTS_TIME_OUT_IN_SECONDS
, which defines the maximum total duration (in seconds) that any single incoming request is allowed to run. This setting helps prevent extremely long-running or stalled requests from hanging indefinitely. It overrides Node.jsâs default 5-minute timeout behavior and ensures more predictable server behavior under heavy or slow traffic. - Now labels from the docker file will take precedante over labels from docker build commands since those are passed from "FROM" commands. So this will allow users to set their own labels and overide existing one from the base docker image
Changesâ
- Helm chart - Increased the default
vulnerabilitiesVolumeSize
value from2Gi
to4Gi
to provide enough space for the new Grype auto-update cron job. You can keep it at2Gi
if updates are disabled. - Helm chart - Added a new
livenessProbe
to the server deployment. This probe checks the/health/full
endpoint and will automatically restart the pod if the server becomes unresponsive. - Helm chart - Added a new
readinessProbe
to the server deployment. This probe checks the/health
endpoint and controls when the server is ready to receive traffic. - Helm chart - Added both
livenessProbe
andreadinessProbe
to the client deployment as well, using the same health check logic.
Make sure your
values.yaml
includes the newlivenessProbe
andreadinessProbe
configurations before upgrading, as both are enabled by default in this version.
- Helm chart & Docker compose - Updated the default nginx image in both Helm and Docker Compose setups to
nginxinc/nginx-unprivileged:1.27.4-bookworm-perl
.