Replies: 1 comment 5 replies
-
|
either Debian has to update to a more recent version >= 2.4.15.2 of mod_auth_openidc, or they have to include a patch to avoid this crash (see: c2f200f) or you can download a newer version manually from the Assets section of the Releases page, or you can compile a more recent version manually |
Beta Was this translation helpful? Give feedback.
5 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
Hi, everyone! We've been upgrading our Apache Docker image to php:8.3-apache, which uses Apache/2.4.59 (Debian). We are adding the latest libapache2-mod-auth-openidc version found in the debian repo, currently 2.4.12.3-2+deb12u1.
As the image is running in an Kubernetes deployment behind a reverse proxy, we've set the
OIDCXForwardedHeadersheader, which assures that the redirect_uri is calculated correctly. So far, everything works fine.As per Kubernetes best practise, we also configure Apache to serve a static endpoint that can be used to query if it is still alive - in our case /live. This endpoint is meant to be regularly queried by the cluster, and unless 200 is returned, the pod (=process) is restarted.
Of course, this endpoint does not need to be SSO protected, so it is served by a
<Location>that does not specifyAuthType openid-connect. At the same time, since the request is coming from inside the cluster, there are not Forward headers set.It turns out, in this setup, we see the following log output:
This looks like 2 issues to me:
<Location>that das OIDC enabled, IMHO the config check should not be performedIf I disable the requests to the /live endpoint, the Apache is serving requests fine, so it seems to be triggered by this request to a location not configured to perform OIDC...
Beta Was this translation helpful? Give feedback.
All reactions