Skip to content

docs: Add Debian 12 breaking change for opendmarc package#3841

Merged
polarathene merged 3 commits intomasterfrom
docs/changelog-add-deb12-breaking-change-opendmarc
Jan 27, 2024
Merged

docs: Add Debian 12 breaking change for opendmarc package#3841
polarathene merged 3 commits intomasterfrom
docs/changelog-add-deb12-breaking-change-opendmarc

Conversation

@polarathene
Copy link
Copy Markdown
Member

Description

Part of the breaking changes for v14 with Debian 12 package upgrade, this config change was released in 1.4.2, while we previously had 1.4.0. Thus upstream would seem to treat it as a bugfix 👍 (but we should probably still mention this, hence the addition to changelog for awareness)

Can we also get confirmation from @georglauterbach on what the dmarc equivalent behaviour in rspamd has been configured for with DMS? (regardless of explicitly / implicitly). We should probably have this in sync.

Fixes #3828

Type of change

  • Improvement (non-breaking change that does improve existing functionality)
  • Breaking change (fix or feature that would cause existing functionality to not work as expected)
  • This change requires a documentation update (CHANGELOG fix)

Checklist:

  • New and existing unit tests pass locally with my changes
  • I have added information about changes made in this PR to CHANGELOG.md

@polarathene polarathene added area/security kind/improvement Improve an existing feature, configuration file or the documentation area/documentation service/security/dkim-dmarc-spf labels Jan 27, 2024
@polarathene polarathene added this to the v14.0.0 milestone Jan 27, 2024
@polarathene polarathene self-assigned this Jan 27, 2024
Comment thread CHANGELOG.md Outdated
Comment thread CHANGELOG.md Outdated
@georglauterbach
Copy link
Copy Markdown
Member

Can we also get confirmation from @georglauterbach on what the dmarc equivalent behaviour in rspamd has been configured for with DMS? (regardless of explicitly / implicitly). We should probably have this in sync.

DMARC is handled by Rspamd by applying a symbol (with a score) to the result of the DMARC check:

# DMARC
"DMARC_NA" {
weight = 1;
description = "No DMARC record";
groups = ["dmarc"];
}
"DMARC_POLICY_QUARANTINE" {
weight = 1.5;
description = "DMARC quarantine policy";
groups = ["dmarc"];
}
"DMARC_POLICY_REJECT" {
weight = 2;
description = "DMARC reject policy";
groups = ["dmarc"];
}
"DMARC_POLICY_ALLOW" { # no equivalent
weight = -1;
description = "DMARC permit policy";
groups = ["dmarc"];
}
"DMARC_POLICY_ALLOW_WITH_FAILURES" { # no equivalent
weight = -0.5;
description = "DMARC permit policy with DKIM/SPF failure";
groups = ["dmarc"];
}
"DMARC_POLICY_SOFTFAIL" { # == DMARC_POLICY_QUARANTINE
weight = 1.5;
description = "DMARC soft-failed";
groups = ["dmarc"];
}

@polarathene
Copy link
Copy Markdown
Member Author

DMARC is handled by Rspamd by applying a symbol (with a score) to the result of the DMARC check

Soooo... it is like the behaviour OpenDMARC had before, or has now? Or neither?

We had a similar discussion recently for the SA_KILL / SPAMASSASSING_SPAM_TO_INBOX=1 settings where it was possible to prevent rejection or similar actions by always allowing the mail into junk folder.

Is that the same "fix" workaround? reject = 100?

@georglauterbach
Copy link
Copy Markdown
Member

DMARC is handled by Rspamd by applying a symbol (with a score) to the result of the DMARC check

Soooo... it is like the behaviour OpenDMARC had before, or has now? Or neither?

I actually don't know whether you can compare this very well... OpenDMARC acts on a single check and "only" gives a decision to Postfix based on this check IIRC. Rspamd on the other hand handles multiple checks (SPF, DKIM, DMARC, denylists, etc.), and will provide a decision based on these multiple checks - the DMARC check is one of them.

We had a similar discussion recently for the SA_KILL / SPAMASSASSING_SPAM_TO_INBOX=1 settings where it was possible to prevent rejection or similar actions by always allowing the mail into junk folder.

Is that the same "fix" workaround? reject = 100?

Yes, reject = 100; is the way of disabling rejection (only GTUBE will still trigger a rejection). Theoretically, a mail could have a score of >100, but the highest I have ever seen was 18. For anyone who can send me an e-mail that gets a score of >50, I'll buy you a 🍺 haha :D

@polarathene polarathene merged commit 4a05d7b into master Jan 27, 2024
@polarathene polarathene deleted the docs/changelog-add-deb12-breaking-change-opendmarc branch January 27, 2024 21:23
@ghnp5
Copy link
Copy Markdown
Contributor

ghnp5 commented Jan 29, 2024

Hey

I didn't even know about postfix's "hold" queue!

Just checked, and I have loads of messages on hold.
I see they're all spam, so I didn't lose much.

Do you know if it's possible to make quarantined emails to either go to a specific folder (like Spam/Junk), or a separate mailbox?

It seems to be hard to see the emails on "hold", as we have to use postqueue -p and then postcat, etc.

Thanks!!

@polarathene
Copy link
Copy Markdown
Member Author

Do you know if it's possible to make quarantined emails to either go to a specific folder (like Spam/Junk), or a separate mailbox?

I am not familiar with working with Postfix hold queue, I know you can check and manipulate the queue but not sure about redirecting that mail.

If the quarantined mail has specific information / metadata in the mail headers that can be matched then a sieve script could probably handle that like we do with redirecting spam delivered to junk folder by default.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

area/documentation area/security kind/improvement Improve an existing feature, configuration file or the documentation service/security/dkim-dmarc-spf

Projects

None yet

Development

Successfully merging this pull request may close these issues.

v14.0.0: with switching to debian:12 bookworm opendmarc behaviour changes

3 participants