Just for fun: Now using BigQuery to import my credit card bills and analyze them, lol. (And the higest went to Uber Eats with no competition…)

SELECT
  shop,
  FORMAT_DATE('%Y-%m', date) ym,
  SUM(amount) as `count`
FROM `credit_csvs.utf-8`
GROUP BY ym, shop
ORDER BY count DESC

Replacing my Python analyzing script so I can just focus on adding more complex queries for future uses.

Interesting findings that for emails sent to Proofpoint protected MX, they might get rejected if the source domain has a stronger DMARC policy, e.g. strict DKIM alignment.

https://help.proofpoint.com/Proofpoint_Essentials/Email_Security/Administrator_Topics/Other_Features/Why_does_DKIM_fail

It all begins with a DMARC report I received from Outlook…

<policy_evaluated>
<disposition>none</disposition>
<dkim>fail</dkim>
<spf>pass</spf>
</policy_evaluated>

I was then eager to find out why DKIM failed in this case, and did some tests with a domain that has its MX pointed to pphosted servers. In the end, I was able to reproduce the issue with a URL link in my email so that it will be rewritten to URL Defense link, as a result, the email body was modified, and resulted in dkim=fail (body hash did not verify).