[bug]: Many TLS errors

[bloodchen]

What happened?

Hi,

From the message Q, we found many failed emails with a common error "TLS error from 'extmx.eyou.com': Handshake failed: received fatal alert: HandshakeFailure" . The domain names (extmx.eyou.com) are different but the rest error message are the same. We have set up STARTTLS to optional and it works for most of the time.

How to make it work for all email servers?

Thanks

Jeff

How can we reproduce the problem?

I can reproduce the problem by doing the following steps:

Version

v0.9.x

What database are you using?

RocksDB

What blob storage are you using?

RocksDB

Where is your directory located?

Internal

What operating system are you using?

Docker

Relevant log output

received fatal alert: HandshakeFailure", elapsed = 636ms
	
Connecting to remote server (delivery.connect) queueId = 195073613995485124, from = "[email protected]", size = 1167, count = 1, domain = "sapo.pt", hostname = "mx.ptmail.sapo.pt", localIp = 0.0.0.0, remoteIp = 212.55.154.36, remotePort = 25, elapsed = 194ms

TLSA record not found (dane.tlsa-record-not-found) queueId = 195073613995485124, from = "[email protected]", size = 1167, count = 1, domain = "sapo.pt", hostname = "mx.ptmail.sapo.pt", strict = false, elapsed = 163ms
Sat, 07 Sep 2024 07:56:48	

Error fetching MTA-STS policy (mta-sts.policy-fetch-error) queueId = 195073613995485124, from = "[email protected]", size = 1167, count = 1, domain = "sapo.pt", causedBy = Invalid DNS record type (mail-auth.dns-invalid-record-type), strict = false, elapsed = 155ms
Sat, 07 Sep 2024 07:56:47	

Error fetching TLS-RPT record (tls-rpt.record-fetch-error) queueId = 195073613995485124, from = "[email protected]", size = 1167, count = 1, domain = "sapo.pt", causedBy = Invalid DNS record type (mail-auth.dns-invalid-record-type), elapsed = 156ms
Sat, 07 Sep 2024 07:56:47	

New delivery attempt for domain (delivery.domain-delivery-start) queueId = 195073613995485124, from = "[email protected]", size = 1167, count = 1, domain = "sapo.pt", count = 56

Code of Conduct

• I agree to follow this project's Code of Conduct

[mdecimus]

Hi,

All those servers that Stalwart can't successfully connect to over TLS are using insecure cipher suites which are not and will not be supported by rustls (the TLS library used by Stalwart). You need to contact the remote server's administrator so they upgrade their systems or disable TLS for these hosts.

[bloodchen]

Well, it's impossible to contact each mail server . Why not try the non TLS connection automatically when TLS connection fail? ( since I choose STARTTLS to be optional) And is there a config we could specify which servers shall not use TLS?

[mdecimus]

That is possible, see this https://stalw.art/docs/smtp/outbound/tls/#handling-tls-errors