Spam/Not Spam seemingly does nothing

[amnesia]

How long does it take the learning mechanism to learn or unlearn spam?

I have one piece of spam I get every day, sometimes 2 or 3 times, it's always very similar messages. I have marked probably 200 of them as spam, and I still get them in my inbox.

Conversely, I have admin emails being forwarded to me, and the daily mail report always ends up in my junk folder, despite every day marking it as Not Spam. How long will it take to learn?

[uxbod]

Please right click on one of those emails and select Show Original. When that opens please post the lines that start with X-Spam-

[amnesia]

From the spam:

X-Spam-Flag: NO
X-Spam-Score: -3.315
X-Spam-Level:
X-Spam-Status: No, score=-3.315 tagged_above=-10 required=6.6
tests=[AM:BOOST=-3, BAYES_00=-1.9, DATE_IN_PAST_03_06=1.592,
DKIM_ADSP_CUSTOM_MED=0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1,
FREEMAIL_FROM=0.001, HTML_MESSAGE=0.001, RCVD_IN_DNSWL_NONE=-0.0001,
T_RP_MATCHES_RCVD=-0.01] autolearn=no

It's spam from a yahoogroups.com list, which I never subscribed to so when I try to unsubscribe it fails. I'm guessing that the domain key makes it instantly valid?

This is the daily mail report:

X-Spam-Flag: YES
X-Spam-Score: 7.852
X-Spam-Level: *******
X-Spam-Status: Yes, score=7.852 tagged_above=-10 required=6.6
tests=[ALL_TRUSTED=-1, BAYES_99=3.5, IP_LINK_PLUS=0.012,
NORMAL_HTTP_TO_IP=0.001, NUMERIC_HTTP_ADDR=1.242, T_FRT_CONTACT=0.01,
T_RP_MATCHES_RCVD=-0.01, URI_HEX=1.122, URI_NOVOWEL=0.5,
URI_OBFU_WWW=2.475] autolearn=no

I'm not sure why it's not working since it is coming from itself.

[uxbod]

For the daily mail report what is the output from

su - zimbra
zmprov gs `zmhostname` zimbraMtaMyNetworks

[amnesia]

zimbra@zimbra:~$ zmprov gs `zmhostname` zimbraMtaMyNetworks
# name zimbra.polarisusa.com
zimbraMtaMyNetworks: 127.0.0.0/8 10.0.1.0/24 209.195.9.0/28

All of which looks correct to me...

[uxbod]

Bug 53060 - Local mails marked as spam

[amnesia]

Ah thanks. I agree, the daily mail report generated by the server itself should bypass the spam check.