3 Replies Latest reply on Oct 30, 2017 9:54 PM by andy777

    ESM Web Filter Policies

    r_gine

      We've got a BlueCoat ProxySG in our environment where we're utilizing URL categorization for content filtering, etc.

       

      I see there are a lot of "out of the box" policy ACE rules designed to detect policy violations (etc, pornography, gaming, etc) and these rules utilize web-filtering normalized events.  I'm trying to understand how I would set up my proxy log data to be normalized based on the URL category as the default BlueCoat parsers do not seem to normalize data based on the URL category.

       

      Any thoughts on how I might accomplish this?

        • 1. Re: ESM Web Filter Policies
          andy777

          When an out of the box rule doesn't match your use case don't hesitate to modify it or make one that does. Some shops turn off all out of the box content and only enable/create rules that map to their use cases. There are a couple of approaches depending upon your goal. If you want to be notified of particular events, just create a Field Match alarm; it includes Boolean logic so it can serve in place of many simple correlation rules. It also fires much faster since the alarm fires when the log hits the Receiver.

           

          If you are more concerned about the Normalization categories, for reporting or to trigger some other rule, you could create a rule for each category that you were interested in which included the correct Normalization

          I would look at the list of categories and depending upon how granular I wanted to be I would either create a rule for each category or group of categories I was interested in.

           

          bc-category.PNG

           

          or if you're feeling adventurous, you could copy the ProxySG parsing rule to a custom rule and hard code the category and desired normalization for each one that you want to break out so it would normalize the way that you want via parsing instead of correlation.

          • 2. Re: ESM Web Filter Policies
            r_gine

            Andy -- I tried to take the 'Copy the ProxySG parsing rule' approach because I want the event to be normalized during parsing. Ultimately I want to build some use cases that would require some base lining and deviation.

             

            I want to normalize the following events as they are coming into the receiver based on the category of the BlueCoat event:

             

            <Malware Monitoring Use Case>

            Malicious Sources/Malnets

            Malicious Outbound Data/Botnets

            Phishing

             

            <Corporate Acceptable Use Policy Violation/Legal Liability Use Case>

            Pornography

            Nudity

            Child Pornography

             

            <Potentially Malicious Insider Use Case>

            Software Downloads

            Hacking

            Remote Access Tools

            Proxy Anonymizers

             

            I copy/pasted the ProxySG Access Log parser (which we have modified to include all of the useful fields left out of the OOTB parser) and simply added content-strings for the following categories. These have been in place for a week w/ no "matches" or events parsed from those parsers. I can query ESM Category = Remote Access Tools and see it is not parsing w/ the expected 'Remote Access Tools' parser. Not sure what the issue is...

             

            regex in 'Parsing' tab confirmed 'Good'.

             

            1) Are actions set appropriately?

            2) If I move from the Data source view to the 'Default Policy' should these parsers be enabled or disabled? (They're currently enabled while the OOTB Proxy parsers are disabled.)

             

             

             

            • 3. Re: ESM Web Filter Policies
              andy777

              I like your approach. Using the Content Strings is a good idea. I think where you could be getting stuck is that there are auto-learned rules that need to be delete before your custom rules are used to create new auto-learned rules.

               

              You can delete auto-learned rules under Data Source below ASP on the left tree. Delete and reroll your policy for your BCs.