We are using open-idconnect with apache-proxy for login in sonarqube.
If a user with “umlaut” like “ö” loggs in via oidc, the ö is shown in the users list as ö.
If i create a user locally in sonarqube. The format is correct “ö”.
If i make a tcpdump -X between apache-proxy and sonarqube i can see the ö is just “…” (two dots).
A strace on the apache-proxy (oidc-provider <–> apache) shows “\303\266” where the ö is.
We are using Version 8.9.10 (build 61524) Community Edition
Hey there.
If you’re using GitHub - vaulttec/sonar-auth-oidc: OpenID Connect (OIDC) Plugin for SonarQube, I would recommend opening an issue with the maintainer on the GitHub repo!
Hey, thanks for fast reply!
I made another tcpdump on localhost:9000 (apache <–> sonar)
I put the Hex from the TCP-Packet (with the “ö”) into an Hex > Ascii-Converter:
656f c3b6 6c72
in ASCII:
eo ö lr
Now i convert “c3 b6” into UTF-8 its “ö”.
Alle theese hex-values are ASCII and only the “c3 b6” is UTF-8.
How can i tell sonar to encode this “c3 b6” in UTF8 ?
Hey there.
You seem to have… completely ignored my reply. 
If you’re authenticating with OpenID via the plugin mentioned, you should raise an issue on their GitHub repo. That’s the side that handles receiving the name/username
Sonarqube is the receiving part here (not the sender…). Anyway, will create an issue there.
Thanks.
Okay now i switched the charset in the proxy to latin1. Now the “ö” is correct encoded in sonarqube.
But it is possible to change the default system encoding to utf-8 instead of latin1 ?
We have many systems using the proxy-settings and have to create a seperate config only for sonar.