Social Login's Authorized redirect URIs are using http instead of https and other errors

[maovidal]

Information about bug

The guide at: https://docs.erpnext.com/docs/user/manual/en/google_settings#how-to-set-up-google-settings indicates the Authorized JavaScript origins to set at API & Services > Credentials in Google.

However, ae least for the Social Login, ERPNext is using:
http://*redacted*/api/method/frappe.integrations.oauth2_logins.login_via_google. This error is informed by Google when requesting the error detail at the popup that follows the expected user login.

Note the http instead https in the origin, and that this string is not o the documentation.
Using temporally http works.

There is also other malformed strings, like the one sent by the contacts integration.

Module

integrations

Version

ERPNext: v15.11.1
Frappe Framework: v15.12.0

Installation method

docker

Relevant log output / Stack trace / Full Error Message.

No response

[kholidraf]

thank you for reporting the bug. Today I just tried to fresh-install the bench and erpnext,. I did integrate the google contact and the problem still occurred. So it seem to be a bug for V15.

I've put this bug here too: https://discuss.frappe.io/t/google-drive-calendar-integration-works-but-contact-failed/116434

[ealecho]

facing the same issue

[ealecho]

facing the same issue

So i was able to resolve the issue by adding the following to my site-config.json

“restart_supervisor_on_update”: true, “restart_systemd_on_update”: true,

and if you use docker like me, add the following nearer the start of your init.sh
bench --site {{site-name}} set-config restart_supervisor_on_update 1 bench --site {{site-name}} set-config restart_systemd_on_update 1

Thanks, @revant. This post was helpful. https://discuss.frappe.io/t/frappe-docker-2023-production-setup-password-reset-email-link-added-with-port-8000/101844/9

[kholidraf]

facing the same issue

So i was able to resolve the issue by adding the following to my site-config.json

“restart_supervisor_on_update”: true, “restart_systemd_on_update”: true,

and if you use docker like me, add the following nearer the start of your init.sh bench --site {{site-name}} set-config restart_supervisor_on_update 1 bench --site {{site-name}} set-config restart_systemd_on_update 1

Thanks, @revant. This post was helpful. https://discuss.frappe.io/t/frappe-docker-2023-production-setup-password-reset-email-link-added-with-port-8000/101844/9

Thanks for the hints.

I didn't use docker for running bench. Is it ok to edit init.sh directly?

[ealecho]

I edited the init.sh file directly and it worked

[notz]

The same issue happens with official kubernetes helm chart.

Setting the host_name variable helps.

bench --site {{site-name}} set-config host_name https://{{site-name}}

[notz]

The problem is in frappe docker image, because it doesn't work with upfront SSL termination (ingress controller). See: frappe/frappe_docker#1381