Plain text files get blocked on content filter

[Integration Support]

Hey all,

I'm encountering an issue with the content filter I configured in ESA. I've set it to only allow specific file types, but sometimes plain text emails without attachments are being incorrectly identified as "unidentified files" and sent to quarantine.

I've attached an example of the rule I configured and mails that I download from the quarantine 1 HTML format and Plain text format  for your reference.

Thanks,

[Ken Stieers]

In the condition section, at the top left, you have it set for "If one or more conditions match"
And your first condition is "attachment-filetype != "mp4"...
So all attachments that aren't MP4 will get quarantined. The engine starts at the top, and if one matches, it quits checking.

What are you trying to do?

[Integration Support]

Thank you for your response.

What I'm trying to achieve is to create a whitelist for allowed file types. Any files not on this list should be sent to quarantine.

The image I uploaded was from a lab environment. My question aims to understand what syntax to use to prevent emails generated in plain text format, which the system identifies as an attachment "without a type," from being blocked.

[Dustin Anderson]

you'll need to make it a match all rule then, but then would say like attachment is not (allowed types), Attachment does not contain \.

it does regex and the \ is an escape character so if the attahcment doesn't contain a period it's let through. I would also from experience allow message.htm and message.html through as some email systems send HTML messages as attachments.

Keep in mind if you quarantine if the email has 1 of the not allowed types it'll quarantine it all. So you could look at sending a copy to a quarantine and stripping the unwanted attachments instead. 

[Dustin Anderson]

As Ken stated, you have it set to any match, and a basic condition of the attachment not being MP4, so everything else will match and quarantine. Best it to set to match on what you don't want for less false positives. Below is an example of what we had.