| Products & Services |
 |
Web Site Hosting |
 |
Web Site Hosting
|
|
Dedicated Servers |
|
Dedicated Servers
|
|
Colocation |
|
Colocation
|
|
Email Solutions |
|
Email Solutions
|
|
Domain Registration |
|
Domain Registration
|
 |
|
| It is not necessary to completely understand how header tags work or how they are determined in order to build an effective content filters. Content filters are simple rules that direct the mail delivery system to deliver mail, or reject mail based on simple comparisons.
Content filters are not only specific to filtering SPAM. Content filters can be used to direct normal mail to different accounts or folder locations.
|
In order to understand how to effectively control spam you must first understand the concept of content filtering. The final stage of email delivery is a very simple process. The mail server takes the inbound message that has been through all it's checks and places it into the inbox of your domain mail account where it waits for you to read it.
Content filters sit between the final delivery of the message and your inbox. They are simple plain english rules that you construct which can alter the destination of the email delivery. Content filters don't only have to be used to filter out SPAM, they can be used as organizational tools. For example, your regularly receive email from mymom@aol.com and you don't want that going into your inbox with all the other mail you could construct a simple rule that looks at the incoming email and if it determines it's from mymom@aol.com it can direct the email into a special folder you created under your account called "mom".
For the purpose of this tutorial we will use content filters to look at the spam tags in the header of the emails we receive and try to eliminate most if not all of the unwanted mail.
One more thing you need to know about content filters and this is a critical point. Content filters can be defined on a domain level, or on a local user level. What that means is that your domain administrator can log into the email administrator account and create a content filter that will affect every user in the entire domain. The other very important thing to remember is that content filter rules are process in order so your administrators rules will be processed before your rules are processed.
Here is an example to help understand rule processing. Again you have a rule that tells the system to look for mail from mymom@aol.com and put it into the "mom" mail folder. But all of your mail from mom never reaches you. You check with your system administrator and find out that they have set a rule that directs the server to delete all mail from mymom@aol.com. His rule gets processed before your rule does so you never see the mail. This is normally not a problem but is something worth knowing.
|
| In short, there are no rules. In the following examples we will illustrate what we have found to be the most effective way to use content rules and it involves both domain content rules, and local user rules. But this is not the "law" when it comes to content rules. You can apply all the content rules to the entire domain, or you can apply all the content rules to each and every mail account. The choice wil be yours. There is however one thing that should always be done. Whenever you build a content filter rule, either on a domain level or a user level or both, you should ALWAYS make your first rule a "White List" rule.
What is a white list rule. A white list rule trumps all other rules as long as it is on top of the list of rules. As we said before rules are processed in order. A white list rule will match inbound mail against a "trusted senders" list you can manage and if it finds that the inbound traffic matches any entry in the trusted senders list it will process that rule and not any other rule.
A white list example. Lets say that you are filtering all mail on a spam level of "Medium" (we will get to that soon) and your client is in Korea and his mail is tagged as medium spam. As soon as his mail hits that content rule his mail will go to the spam bucket rather than be delivered. But if your domain or user account has a white list rule as the first rule and his email address or domain is listed in that white list, his mail will go to the inbox and not be directed to the spam bucket because no other rules will be applied to his account.
Now as we said before you can apply all the rules at the domain level and not bother the individual user with the need to have any content filters. That works very well and there is nothing wrong with that. But you must keep in mind that the white list administration now belongs to the email administrator so if you find one of your emails going to spam you will need to ask your email administrator to add the domain or email address to the white list.
If you perform all the content filtering on the local user level you have complete control over your own white list (trusted senders) and that works well for some people. But if your organization has hundreds of employees you will need to see that they all set their content filters.
Somewhere in the middle there is a much easier way to manage spam. What we have adopted is a two tier content filtering scheme. Both the domain administrator and the local user each have their first rule as a white list rule. The domain adiminstrator has two additional rules which detect spam at the 99.9% level and they simply delete the SPAM. So already you have reduced the amount of spam hitting the local user account by a significant amount. The local user has their own white list rule and only one other rule which focuses on a single spam identifier. This identifier may only identify spam wtih a 80% confidence level and may have some false positive results. But the local user can then manage their own white list without the intervention of the domain administrator.
|
|
Cart