Closed anoadragon453 closed 5 years ago
then the certificate gets marked as valid even if
example.com
is included
this sounds like correct behaviour?
people should move to .well-known.
The idea is that a .well-known
is normally unnecessary and you should just use the right cert
LetsEncrypt doesn't seem to set up an IP address into the cert by default though.
Indeed, certificates don't usually include an IP address, but that shouldn't matter as we should check for the original server name, not the target.
then the certificate gets marked as valid even if example.com is included
Did you mean invalid?
There is a separate problem with IP addresses: I don't think it is valid for a SRV to point to an IP address. You might want to check that though.
@anoadragon453 I'm still very confused about what this issue is talking about.
This issue spawned as a configuration error as part of @benparsons' server, which afaik used the community Ansible playbook (which may need to be adjusted to not set the SRV record to an IP address).
As SRV cannot be an IP address, this makes this whole issue moot.
The matrix-docker-ansible-deploy playbook doesn't set SRV records on people's behalf.
In its Configuring DNS, we do advise people to set a SRV record (before Synapse v1.0, at least), but it references a matrix.<domain>
host, never an IP address.
So.. it's interesting exactly what happened and how.
so, no solution was found for this issue.
If a server is using an SRV record that leads to an IP address and that IP address is not included in the target server's federation TLS certificate, then the certificate gets marked as valid even if
example.com
is included (it's a case of trying and failing to validate xx.xx.xx.xx:8448 vs. example.com:8448).I'm of the mind that we should still fail, but use the new errors output to say that this is the case and that people should move to .well-known.