All,
As mentiomed over on Mailop, I've come up with some code to poll the Umbrella
1m (plus the root zone and the public suffix list) for adoption of t= as well
as pct=, so that administrators can determine when they feel ready to tell
their dmarc-checkers to stop honoring the pct= field.
I'm choosing to leave parsing of pct= in-place in the next release of OpenDMARC
but make it user-configurable, to not violate POLA.
In the top 10,000 u1m domains, I found about 30 with pct=0, including several
major corporate ones (ciscospark.com, several shopify domains) and a NY state
university (sunysb.edu) that have non-p=none policies. Is this a small amount?
Probably. But it's nonzero.
My polling scripts are in OpenDMARC/contrib on the develop branch. I'm
planning to put up a website for this at some point shortly with pretty graphs
to track this over time.
Just because I find it interesting and I'm already running the DNS queries, I
also report on:
* Domains with pct= where it would not make sense.
* How many RUA's post third-party consent records per 9990 section 4 (or 7489
SS 7.1)
* How many domains in the u1m have dmarc records further up in the parent
(presence in the u1m is obviously not an indicator that a domain originates
mail).
Domains queried : 1000003
Errors : 7146
Have DMARC : 110676 (11.1%)
Have pct= : 33737 (30.5% of DMARC)
Have psd= : 19 (0.0% of DMARC)
psd=y : 3
psd=n : 16
Have t= : 168 (0.2% of DMARC)
t=y : 25
t=n : 143
p=reject/quarantine with pct=100 (no-op) : 22647
p=none with pct= (no effect) : 8323
RFC 7489 S 7.1 external reporting consent (dmarc-pct-survey-2026-06-25.tsv):
Pairs checked : 74346
Domains w/ unconsented rua= : 9511
Domains w/ unconsented ruf= : 5067
Parent-label walk (dmarc-pct-survey-2026-06-25-extra.tsv):
Unlisted parents checked : 2101
Unlisted parents w/DMARC : 163
(unlisted in this case means "is not in the u1m")
-Dan
_______________________________________________
dmarc mailing list -- [email protected]
To unsubscribe send an email to [email protected]