OIDC Error with Openwebui Latest Update 0.6.11

[firmansi]

After recent update, when login with OIDC, the authorization directed to localhost:3000m which is quite odd since I have this instance running on AWS with public domain. But then I made a change in Admin Panel at "Setting > General" and change the url of the OpenWebui to the real one (my public domain)

I think someone need to fix this issue, since it didn't happn in previous updates, and to someone who barely unfamiliar with OpenWebUI if they face this issue, they might not have the answer in Cloudron Documentation. And by the way, this issue not exist in previous versions

I show below the part in OpenWebUI that needs to be changed if this problem persist to login with OIDC

[james]

Hello @firmansi I will look into it.

[james]

Just did a fresh installation of the Open Web UI Cloudron app with the latest version.
When viewing the mentioned section I can see a correctly configured WebUI URL.

@firmansi please restart the app once.
After that, go into the web terminal of the app and run the following commands and provide the output here.

echo $CLOUDRON_OIDC_ISSUER

and

echo $CLOUDRON_APP_ORIGIN

Also, please check your /app/data/env.sh file and provide the content of that file with redacted credentials.

[firmansi]

@james when I did the echo command, all is normal, I mean it goes to my normal domain, as I said that I have changed the setting in Admin Panel OpenWebui
Here is the env configuration

# See https://docs.openwebui.com/getting-started/advanced-topics/env-configuration/ for more options
# Change this to false to disable local ollama and use your own
export LOCAL_OLLAMA_ENABLED=false

export DEFAULT_USER_ROLE="user"


export STATIC_DIR="/app/data/image_static"

# When using remote ollama, change this to the ollama's base url
# export OLLAMA_BASE_URL="https://<remote-ollama>"

# When local ollama is enabled, this is location for the downloaded models.
# If the path is under /app/data, models will be backed up. Note that models
# can be very large. To skip backup of models, move the models to a volume (https://docs.cloudron.io/volumes/)
# export OLLAMA_MODELS=/app/data/ollama-home/models

[james]

Thanks, @firmansi
I've already found the issue and an app update is imminent.

[firmansi]

@james After recent patch to latest update 0.6.11 I got another issue now the OpenwebUI is unable to start with error message as follow

"May 28 21:24:23 => Healtheck error: Error: connect EHOSTUNREACH 172.18.20.162:8080"

Please advise what should I do ?

[nebulon]

If you restart the app, what other errors during startup do you see in the logs?

[firmansi]

@nebulon Same message as mentioned, I have restarted 4x including restart the server

[joseph]

There are probably more messages in the logs. Healthcheck error does not come from the app itself. If you keep log window open and restart the app, you should see a better error message.

[firmansi]

@joseph here the other message on top of the message i've shown you

May 28 22:26:40 box:shell services /usr/bin/sudo -S /home/yellowtent/box/src/scripts/mkdirvolume.sh /home/yellowtent/appsdata/1b891327-e698-4363-999c-312b7cb534f3/data
May 28 22:26:40 box:services Setting up postgresql
May 28 22:26:40 box:services setupAddons: setting up addon postgresql with options {}
May 28 22:26:40 box:services Setting postgresql addon config to [{"name":"CLOUDRON_POSTGRESQL_URL","value":"postgres://user1b891327e6984363999c312b7cb534f3:ee56905ff986b695a223cebd85f4cf14d853a209a4b9d52deba2115c24aec9ae04ca7c6d9642397edbfbf8a155f4d174823bd57efbfe2853a1cc908d0e31ec29@postgresql/db1b891327e6984363999c312b7cb534f3"},{"name":"CLOUDRON_POSTGRESQL_USERNAME","value":"user1b891327e6984363999c312b7cb534f3"},{"name":"CLOUDRON_POSTGRESQL_PASSWORD","value":"ee56905ff986b695a223cebd85f4cf14d853a209a4b9d52deba2115c24aec9ae04ca7c6d9642397edbfbf8a155f4d174823bd57efbfe2853a1cc908d0e31ec29"},{"name":"CLOUDRON_POSTGRESQL_HOST","value":"postgresql"},{"name":"CLOUDRON_POSTGRESQL_PORT","value":"5432"},{"name":"CLOUDRON_POSTGRESQL_DATABASE","value":"db1b891327e6984363999c312b7cb534f3"}]
May 28 22:26:40 box:services Setting up OpenID connect
May 28 22:26:40 box:services setupAddons: setting up addon oidc with options {"loginRedirectUri":"/oauth/oidc/callback"}
May 28 22:26:40 box:tasks update 1335: {"percent":70,"message":"Creating container"}
May 28 22:26:40 box:apptask createContainer: creating container
May 28 22:27:23 box:shell apptask /usr/bin/sudo -S /home/yellowtent/box/src/scripts/configurelogrotate.sh add 1b891327-e698-4363-999c-312b7cb534f3 /tmp/1b891327-e698-4363-999c-312b7cb534f3.logrotate
May 28 22:27:23 box:apptask startApp: starting container
May 28 22:27:24 box:tasks update 1335: {"percent":90,"message":"Configuring reverse proxy"}
May 28 22:27:24 box:tasks update 1335: {"percent":100,"message":"Done"}
May 28 22:27:24 box:tasks setCompleted - 1335: {"result":null,"error":null}
May 28 22:27:24 box:tasks update 1335: {"percent":100,"result":null,"error":null}
May 28 22:27:24 box:taskworker Task took 536.239 seconds

[firmansi]

@joseph and the other log is

May 28 22:32:20 => Copying static files
May 28 22:32:20 => Ensure directories
May 28 22:32:21 => Enabling local ollama
May 28 22:32:21 => Loading configuration
May 28 22:32:21 => Setting permissions
May 28 22:32:21 ERROR: syntax error at or near "m"
May 28 22:32:21 LINE 1: ... }, { "title": [ "Explain options trading", "if I'm familiar...
May 28 22:32:21 ^
May 28 22:32:27 => Healtheck error: Error: Timeout of 7000ms exceeded
May 28 22:32:37 => Healtheck error: Error: Timeout of 7000ms exceeded
May 28 22:32:47 => Healtheck error: Error: Timeout of 7000ms exceeded
May 28 22:32:57 => Healtheck error: Error: Timeout of 7000ms exceeded
May 28 22:33:07 => Healtheck error: Error: Timeout of 7000ms exceeded
May 28 22:33:17 => Healtheck error: Error: Timeout of 7000ms exceeded
May 28 22:33:21 => Copying static files
May 28 22:33:21 => Ensure directories
May 28 22:33:22 => Enabling local ollama
May 28 22:33:22 => Loading configuration
May 28 22:33:22 => Setting permissions
May 28 22:33:22 ERROR: syntax error at or near "m"
May 28 22:33:22 LINE 1: ... }, { "title": [ "Explain options trading", "if I'm familiar...
May 28 22:33:22 ^
May 28 22:33:27 => Healtheck error: Error: Timeout of 7000ms exceeded
May 28 22:33:37 => Healtheck error: Error: Timeout of 7000ms exceeded
May 28 22:33:47 => Healtheck error: Error: Timeout of 7000ms exceeded
May 28 22:33:57 => Healtheck error: Error: Timeout of 7000ms exceeded

[firmansi]

@joseph I think I am gonna go back to release before this last patch, at least the service is still running eventhough need to enter the value in Administrator Panel with the domain name that runs the openwebui

[ApplegateR]

I have same issues,

3 => Healtheck error: Error: connect EHOSTUNREACH 172.18.18.16:8080
May 28 09:24:03 => Ensure directories
May 28 09:24:03 => Copying static files
May 28 09:24:03 => Loading configuration
May 28 09:24:03 => Enabling local ollama
May 28 09:24:03 => Setting permissions
May 28 09:24:03 ERROR: syntax error at or near "ll"
May 28 09:24:03 LINE 1: ...ite a sentence for me to fill in the blank, and I'll try to ...

[nebulon]

Oh looks like the start.sh has some shell script escaping error. Not sure why our integration tests worked fine, but we have to investigate. I have revoked the last version for a start until this is resolved. Everyone who got updated already, please revert to the version prior the update.

[ApplegateR]

Thank you

[james]

Hey @ApplegateR and @firmansi
Do either of you two still have the defective version installed?
If yes, please put the app in debug mode.
Open the web terminal and edit the /app/pkg/start.sh.
In line 2 make set -eu to set -eux and run the start script with:

/app/pkg/start.sh

If it fails, please send the output.

[james]

I think I just understood the issue after pondering the error message.
Trying to reproduce it now.

[james]

Solution found, app might be updated soon and looking forward to get feedback.

[ApplegateR]

@James Hi, I restored to a previous backup version that is currently working. I will wait for the next update and let you know if it resolves the issue.

[firmansi]

@james Cool, looking forward to this, I have reverted back to previous version so I can't test the changes to the two lines in script as you mentioned before