While your environment is booting up, it streams a message indicating the status of the boot phase. If something goes wrong, the status will turn red (in both the status section of the pull request on GitHub, and on your team's Dashboard) and an error message will be displayed. Continue reading to learn more detail about each error message, and what you can do to fix it.

Error status messages

Deployment unreachable. Please check port.
This message occurs when the port exposed in your Dockerfile doesn't match the port you set in your peek.yml file.

Error connecting to docker registry.
Double-check that the required environment variables exist in the context of your CI environment, and that they are set to the right values. The required environment variables differ for each container registry – here they are for ECR, GCR, Docker Hub, and GitHub Packages.

Health check response contains invalid HTTP headers.
Because FeaturePeek runs your code within an iframe, the following HTTP response headers are not allowed:

X-Frame-Options: A value of deny or sameorigin will prevent FeaturePeek from working properly. It's recommended to use the frame-ancestors directive of Content-Security-Policy instead.
Content-Security-Policy: If you're including the frame-ancestors directive, you must whitelist https://dashboard.featurepeek.com as a source – the source cannot include 'self' or 'none'.

If you're still encountering problems, please reach out to support.

Image not found in container registry.
FeaturePeek could log into your container registry using your Docker credentials, but couldn't find an image and/or tag in your registry using the metadata we received from the CI ping. Double-check in your registry that you are indeed pushing an image with the expected name and tag.

Ping timeout. Please check your CI.
You may not have set up your Continuous Integration pipeline correctly. This error gets displayed if FeaturePeek does not receive a ping message from your CI pipeline within one hour of a commit getting pushed to a pull request. Double-check that your CI pipeline successfully ran, built your front-end, and pinged FeaturePeek in less than one hour – it's really useful to check the logs of your CI provider.

Received non-200 HTTP status code.
Most likely, your root route / redirects to another page, causing a 3xx HTTP code. FeaturePeek's health checking mechanism requires that the route returns a 200 so we know that your deployment is operational. If you'd like to change the page that the health check checks, change the value of "Health check path" in your project's settings. You may need to restart your deployment after changing your project's settings.

If your root route isn't intended to redirect, you may have a 4xx or 5xx class HTTP error. Double-check your logs to debug. You may need to set environment variables if you haven't done so already.

Running image is in a crash loop.
Your app is crashing. Double-check your logs to debug. You may need to set environment variables if you haven't done so already.
Was this article helpful?
Cancel
Thank you!