From this commit on, api endpoints reply with a 401 error code and non api
endpoints redirect to /error/40X.
This commit also fixes missing restrictions on /loggedin (the "already logged
in page). This was not a security issue, though.
The change also makes error pages automatically redirect the user after few
seconds based on the referrer or the default_redirection_url if provided in the
configuration.
Warning: The old /verify endpoint of the REST API has moved to /api/verify.
You will need to update your nginx configuration to take this change into
account.
Issuer is customizable in configuration so that a company can set its own name
or website. If not provided, default value is 'authelia.com'.
The username is used as label.
This URL is used when user access the authentication domain without providing
the 'redirect' query parameter. In that case, Authelia does not know
where to redirect the user.
If the parameter is defined, Authelia can redirect the user to a default page
when no redirect parameter is provided.
When user is already authenticated and tries to access the authentication
domain, the "already logged in" page is rendered and it now tells the user he
is to be redirected in few seconds and uses this URL to redirect.
This parameter is optional. If it is not provided, there is only a notification
message at the end of the authentication process, as before, and the user is
not redirected when visiting the authentication domain while already
authenticated.
This timeout will prevent an attacker from using a session that has been
inactive for too long.
This inactivity timeout combined with the timeout before expiration makes a
good combination of security mechanisms to prevent session theft.
If no activity timeout is provided, then the feature is disabled and only
session expiration remains as a protection.
A window of 1 means the token is checked against current time slot T
as well as at time slot T-1 and T+1.
A time slot is 30 seconds by default in Authelia.
Now, /verify can return 401 or 403 depending on the user authentication.
Every public API endpoints and pages return 200 with error message in
JSON body or 401 if the user is not authorized.
This policy makes it complicated for an attacker to know what is the source of
the failure and hide server-side bugs (not returning 500), bugs being potential
threats.
Example 1: The user has validated first factor when accessing a service
protected by basic auth. When he tries to access another service protected
by second factor, he is redirected to second factor step to complete
authentication.
Example 2: The user has already validated second factor. When he access auth
service, he is redirected either to /loggedin page that displays an "already
logged in" page or to the URL provided in the "redirect" query parameter.
The notification message pops up and hide after few seconds.
Sometimes, chrome drivers tries to click on a button that moves due
to the notification message animation and thus miss it.
Previously, string "{0}" was replaced by the user dn in the groups_filter
attributes of the LDAP configuration.
However, if the groups children only have a memberUid attribute, one would
like to use the username instead of the user dn.
Since the user dn can be built from the username, "{0}" is now replaced
by the username instead of the user dn so that an LDAP relying on attribute
'memberUid' can be used.
Client and server now have their own tsconfig so that the transpilation is only
done on the part that is being modified.
It also allows faster transpilation since tests are now excluded from tsconfig.
They are compiled by ts-node during unit tests execution.
One can now plug its own SMTP server to send notifications
for identity validation and password reset requests.
Filesystem has been removed from the template configuration file
since even tests now use mail catcher (the fake webmail) to
retrieve the email and the confirmation link.
ACLs can now be defined by subdomain AND resource using pattern matching
with regular expressions.
It allows a very fine-grained access control to backend resources.
[Note] For using example environmnent, user must update its /etc/hosts with
new subdomains updated in README.
Before this fix, the application was simply crashing during execution
when connection to redis was failing.
Now, it is correctly handled with failing promises and logs have been
enabled to clearly see the problem
Before this fix, the redirection URL was stored in the user session,
but this has a big drawback since user could open several pages in
browser and thus override the redirection URL leading the user to
be incorrectly redirected.