Bot Manager Advanced

Bot Manager Advanced adds an additional layer of security that is dedicated to bot detection and mitigation. It is designed to automatically detect good bots (e.g., search bots) and bad bots, including those that spoof good bots, by analyzing requests and behavior. You may even customize how bad bots are detected and mitigated by defining custom criteria that profiles a bad bot and the action that we will take for that traffic. Bot Manager Advanced is also able to mitigate basic bots by requiring a web browser to resolve a JavaScript challenge before our service will resolve traffic. Finally, it provides actionable near real-time data on detected bots through which you may fine-tune your configuration to reduce false positives.

Bot Manager Advanced is a powerful tool through which you may mitigate undesired bot traffic and prevent them from performing undesired or malicious activity, such as scraping your siteRefers to harvesting data from your site., cardingRefers to the process through which stolen credit cards are authorized., taking over accounts through credential stuffing, spamming your forms, launching DDoS attacks, and committing ad fraud.

How Does It Work?

Bot Manager Advanced inspects each request to determine whether the client:

1. Matches a known good bot (e.g., search bot).
2. Is spoofing a known good bot.

3. Matches a rule. A rule defines the criteria that our service will use to identify a bad bot.

   You may identify bots using:

   • Information derived from the request, such as geolocation, IP address, and the URL path.

   • Our request and behavioral analysis that assigns a bot score to the request that defines our level of confidence that it is a bot.

     You may set actions based off of bot score thresholds.
     
     For example, you may redirect requests whose bot score is between 50 and 80% and block requests whose bot score is greater than 80%.

   • The JA3 fingerprint assigned to the request. A JA3 fingerprint identifies a client using key characteristics from a TLS request. This allows us to classify traffic as a specific bot across various IP addresses and ports.

Key information:

• Your configuration determines how our service will handle the above traffic patterns.

• If a request satisfies multiple criteria, then the above order determines the action that will be applied to it. Specifically, the order of precedence is:

  Known Bots > Spoofed Bots > Bots Identified by a Rule
• Bypass the above bot detection measures by creating an exception for one or more URL(s), user agent(s), JA3 fingerprint(s), or cookie(s).

Actions

The type of actions that can be applied to bot traffic are:

Mode	Description
Alert	Generates an alert. Use this mode to track detected threats through the Bots dashboard without impacting production traffic.
Browser challenge	Sends a browser challenge to the client. The client must solve this challenge within a few seconds. Response The results of the above browser challenge determines what happens next. Solved: If the client is able to solve the challenge, then our CDN serves the requested content. Additionally, a cookie will be added to the user's session. This cookie instructs our CDN to serve content to the user without requiring a browser challenge. Once the cookie expires, new requests for content protected by Bot Manager will once again require the client to solve a challenge. Define the duration for this cookie through the Valid for (in minutes) option. Unsolved: If the client is unable to solve the challenge, then our CDN responds with a new browser challenge. Key information: Solving a challenge requires a JavaScript-enabled client. Users that have disabled JavaScript on their browsing session will be unable to access content protected by browser challenges. We strongly recommend that you avoid applying browser challenges to machine-to-machine interactions. For example, applying browser challenges to API traffic will disrupt your API workflow. The HTTP Status Code option determines the HTTP status code for the response provided to clients that are being served the browser challenge. Setting this option to certain status codes (e.g., 204) may prevent clients that successfully solve a browser challenge from properly displaying your site. You may define a custom payload for the browser challenge by enabling the Custom Browser Challenge Page option and then setting the Browser Challenge Page Template option to the desired payload.
Block	Drops the request and the client will receive a 403 Forbidden response.
Custom Response	Returns a custom response. Response body: Define the payload that will be delivered to the client. This option supports the use of event variables to customize the response. Sample payload for a HTML file: <!DOCTYPE html><html> <head><title>Page Not Found</title></head> <body>Page not found.</body> </html> HTTP status code: Defines the HTTP status code that will be sent to the client. Custom response headers: Defines one or more response headers that will be sent to the client. Define each custom response header on a separate line. Syntax: Name:Value Example: MyCustomHeader: True This option supports the use of event variables to customize the response. All characters, including spaces, defined before or after the colon will be treated as a part of the specified header name or value, respectively.
Redirect	Redirects requests to the specified URL. Key information: The HTTP status code for this response will be a 302 Found. Set the URL option to the full URL to which requests will be redirected. Example: http://cdn.mydomain.com/marketing/busy.html

Bot Manager Configuration

Each rule within a Bot Manager configuration identifies bot traffic. Each rule contains:

• Up to 6 conditions that define request identification criteria.

• A rule ID and message that will be associated with requests identified by this rule.

  Assigning a unique ID and message to each rule makes it easy to identify requests detected as a result of a specific rule.

  A rule ID must be a number between 77,000,000 and 77,999,999.

A Bot Manager configuration may contain up to 10 rules.

Custom Bot Detection

A request must satisfy at least one rule before WAF will consider it bot traffic. There are two types of rules, which are:

• Custom Matches: This type of rule is satisfied when a match is found for each of its conditions. A condition defines what will be matched (i.e., variable), how it will be matched (i.e., operator), and a match value.

  Certain variables match on key-value pairs. If you match on multiple keys within a single variable, WAF will only need to find one of those matches to satisfy that variable.
  
  For example, if you set up a request header variable to match for Authorization and Content-Type, then requests that contain either or both of those headers will satisfy that variable.

• Edgio Reputation DB: This type of rule is satisfied when the client's IP address matches an IP address defined within our reputation database. Our reputation database contains a list of IP addresses known to be used by bots.

Conditions

A condition determines how requests will be identified through variables, operators, match values, and negative matching.

Variables

A variable identifies the request element that WAF will analyze. We support the following request elements:

• ASN: Identifies requests by the Autonomous System Number (ASN) associated with the client's IP address.

  Specify a regular expression to match for multiple ASNs.
  Example: 
  Use the following pattern to match for requests from either 15133 or 14153:
  15133|14153

• Bot score: Identifies requests based off a score that defines our level of confidence that it is a bot. This score is calculated by analyzing the request and its behavior. The range for this score is 0 to 100.

• Country: Identifies requests by the country from which the request originated. Specify the desired country using a country code.

  Specify a regular expression to match for multiple country codes.
  Example: 
  Use the following pattern to match for requests from the United States, Canada, or Mexico:
  US|CA|MX

• IP address: Identify requests by the requester's IP address. Specify a comma-delimited list of the desired IP address(es) using standard IPv4/IPv6 and CIDR notation. Specify a subnet by appending a slash (/) and the desired bit-length of the prefix (e.g., 11.22.33.0/22). Do not specify more than 1,000 IP addresses or IP blocks.

  Example:

  192.0.2.20,203.0.113.0/24,2001:DB8::/32
• Ja3: Identifies requests by the JA3 fingerprint assigned to the request. A JA3 fingerprint identifies a client using key characteristics from a TLS request. This allows us to classify traffic as a specific bot across various IP addresses and ports.

• Request cookies: Match against all or specific cookies.

  • All: Do not specify the desired cookie within this variable and specify the desired cookie value or pattern within the Match value option.

  • Specific Cookies: Define the name of the desired cookie within this variable and specify the desired cookie value or pattern within the Match value option.

    Setting up a cookie variable also allows you to define whether WAF uses a regular expression, a negative match, or both when comparing the value assigned to the variable against cookies. Use a negative match to find requests whose payload does not contain the specified cookie.

• Request header: Match against all or specific request headers.

  • All: Do not specify the desired request header within this variable and specify the desired header value or pattern within the Match value option.

  • Specific Request Headers: Define the name of the desired request header within this variable and specify the desired header value or pattern within the Match value option.

    Setting up a request header variable also allows you to define whether WAF uses a regular expression, a negative match, or both when comparing the value assigned to the variable against request headers. Use a negative match to find requests whose payload does not contain the specified request header.

• Request method: Match against request method (e.g., GET and POST).

• Request query: Match against the request's query string. Specify the desired value or pattern within the Match value option.

• Request URI: Match against the request's URL path and query string. Define a URL path that starts directly after the hostname. Exclude the protocol and hostname when defining this property.

  WAF does not transform edge CNAME URLs to CDN URLs prior to performing this comparison.

  Sample values:

  /marketing?id=123456

  /resources/images

• Request URL path: Match against the request's URL path. Define a URL path that starts directly after the hostname. Exclude the protocol, hostname, and query string when defining this property.

  Our service does not transform edge CNAME URLs to CDN URLs prior to performing this comparison.

  Sample values:

  /marketing

  /resources/images

All variables support the ability to match on the number of times that a request element is found within the request. Set up a variable to match on the number of instances instead of inspecting the element for a specific value or regular expression pattern by marking the Count option.

You may define zero or more keys when setting up variables that match on key-value pairs. WAF must find at least one of the specified keys in the request before that variable will be satisfied. For example, if you set up a request header variable to match for Authorization and Content-Type, then requests that contain either or both of those headers will satisfy that variable.

Operators

An operator determines how WAF will compare a match value against the request element identified by a variable.

• Begins with: A match is found when the request element starts with the specified match value.
• Contains: A match is found when the request element contains the specified match value.
• Ends with:  A match is found when the request element ends with the specified match value.

• Exact match:  A match is found when the request element is an exact match to the specified match value.

  Avoid enabling the Negative match option with the Exact match operator. This configuration will not yield the expected set of matches.

• Regex: A match is found when the request element satisfies the regular expression defined in the match value.

• Value match: A match is found when the request element occurs the exact number of times defined in the match value.

  The Value match operator should only be used when the Count option has been enabled.

Match Value

WAF uses a match value to identify threats.

• Default: By default, WAF compares a match value against the request element identified by a variable (e.g., URL path or a request header's value).
• Count: Enable the Count option on a variable to compare this value against the number of times that the request element identified by a variable (e.g., a specific cookie or request header) occurs within the request.

Example:

This example assumes the following configuration:

We will now examine how the Count option affects comparisons for this configuration.

• Disabled: If the Count option has been disabled on the variable, then WAF will compare the value of the Authentication request header to 1.
• Enabled: If the Count option has been enabled on the variable, then WAF will compare the number of times that the Authentication request header occurred in the request to 1.

The type of comparison that will be performed is determined by the Operator option.

Exceptions

You may exempt traffic from bot detection by URL, user agent, JA3 fingerprint, and cookie.

Key information:

• Define each entry on a separate line.
• URL, user agents, and cookies are regular expressions.
• A JA3 fingerprint identifies a client using key characteristics from a TLS request.

• Our service will only bypass bot detection when it finds an exact match for a JA3 fingerprint exception.

  Use the Bot dashboard to find the JA3 fingerprint that corresponds to a false positive.

Bot Manager Configuration Administration

You may create, modify, and delete Bot Manager configurations.

Key information:

• Administer Bot Manager configuration from the Bot Manager page.
• Apply a Bot Manager configuration to production traffic by adding it to a Security Application Manager configuration. Multiple Security Application Manager configurations may use the same Bot Manager configuration.
• It typically takes less than a minute to apply Bot Manager configuration changes across our entire network.

To create a Bot Manager configuration

1. Navigate to the Bot Manager page. How?From the main menu, navigate to More | Security | WAF Tier | Security Rule Manager | Bot Manager.

2. Click Add Bot Rule.
3. In the Name option, type the unique name by which this Bot Manager configuration will be identified. This name should be sufficiently descriptive to identify it when setting up a Security Application Manager configuration.

4. Optional. Set up a browser challenge, custom response, or redirect that can be applied to known bots, spoofed bots, and bots detected through rules.

   Unlike other actions, alert and block actions do not require configuration before they can be applied to bot traffic.

   1. From the Actions section, select the desired action:

      • Browser challenge: Perform the following steps:

        1. From the Valid for (in minutes) option, type the number of minutes for which our CDN will serve content to a client that solves a browser challenge without requiring an additional browser challenge to be solved. Specify a value between 1 and 1,440 minutes.

        2. From the HTTP status code option, determine the HTTP status code for the response provided to clients that are being served the browser challenge.

           Setting this option to certain status codes (e.g., 204) may prevent clients that successfully solve a browser challenge from properly displaying your site.

        3. Serve a custom browser challenge by enabling the Custom Browser Challenge Page option and then setting the Browser Challenge Page Template option to the desired payload.

      • Custom response: Perform the following steps:

        1. From the Response body option, specify the body of the response that will be sent to clients.
        2. From the HTTP status code option, determine the HTTP status code for the response that will be sent to clients.

        3. From the Custom response headers option, define each desired custom response header on a separate line.

           Example:

           MyCustomHeader: True
      • Redirect: Set the URL option to the full URL to which requests will be redirected.
   2. Repeat the previous step as needed.

5. Perform the following steps to automatically detect known bots:

   1. From the Known bots tab, click + Add known bot.

   2. From the Bot Token option, select whether to apply an action to all known bots (all), a specific bot, or to 200+ bots (other):

      Choose other to apply an action to 200+ known good bots. This option excludes the known bots that are listed within the Bot Token option.

   3. From the Action type option, select the action that will be applied to the known bot(s) selected in the previous step.

   4. Repeat the above steps as needed.

6. The Spoofed Bots section in the Known bots tab determines how to handle traffic spoofing the known bots selected in the previous step. From the Action type option, select the desired action.

   The Spoofed Bots section does not apply to the 200+ known bots defined within the other category.

7. Create rules for identifying bots from the Rules tab.

   1. In the Bot ruleset name option, type a unique name.
   2. Click + Add Rule.

   3. In the Rule type option, select the type of rule that will be created.

      • Custom Matches: This type of rule is satisfied when a match is found for each of its conditions.

        1. In the Name option, type a name that identifies the purpose of this rule.
        2. In the Rule ID option, specify a number between 77,000,000 and 77,999,999.
        3. In the Rule message option, type a brief description for this rule.

        4. A custom matches rule automatically includes a default condition. Modify this condition to determine how WAF will identify requests. From the condition's Variable option, select the request element through which WAF will identify requests.

           Learn more about variables.

        5. Certain variables (e.g., request cookies and request header) match on name and value. If you have selected this type of variable, then perform the following steps:

           1. Click + Add Match.

           2. From the Name option, type the desired name.

              For example, match for requests that contain an Authorization header by setting this option to Authorization.

           3. Optional. Mark the Negative Match option to match for requests that do not contain a matching value for the name defined in the previous step.
           4. If you specified a regular expression in the Name option, then you should mark the Regex Match option.
           5. Optional. Add another match through which this variable can be satisfied by repeating steps a - d.

        6. Optional. Mark the Count option to match by the number of instances that a match is found instead of by inspecting that request element.

           Learn more.

        7. From the Operator option, select an operator that determines how WAF will compare the match value to the request element identified by the above variable.

           Learn more.

        8. In the Match value option, type the value that will be compared against the request element identified by the above variable.
        9. Optional. Mark the Negative Match option to match for requests that do not contain a matching value for the value defined in step vi.
        10. Optional. Click + Add Condition to add another condition that must be met prior to request identification. Repeat steps i - ix for this new condition.
      • Edgecast Reputation DB: This type of rule is satisfied when the client's IP address matches an IP address within our reputation database. Proceed to the next step.
8. Optional. Click + Add Rule to add another rule through which WAF may identify bots. Repeat step 7.3.

9. Optional. Identify traffic that will bypass bot detection.

   1. Click the Exceptions tab.

   2. Add the desired URL(s), user agent(s), JA3 fingerprint(s), and cookie(s) as exception(s).

      Place each entry on a separate line.

10. Click Save.

To modify a Bot Manager configuration

1. Navigate to the Bot Manager page. How?From the main menu, navigate to More | Security | WAF Tier | Security Rule Manager | Bot Manager.

2. Click on the desired Bot Manager configuration.

3. Make the desired changes.

   Key tasks:

   • Add, modify, or delete an action.
   • Set or modify the action that will be taken for all known bots, specific known bots, and spoofed bots from the Known Bots tab.
   • Update your exceptions to bot detection by adding, modifying, or deleting entries from the Exceptions tab.
   • Change the type of rule from the Rule type option.

   • Custom matches only

     Delete variables and matches within a variable by clicking .

   • Custom matches only

     Delete a condition by clicking Delete Condition.

     A rule must have at least one condition. Therefore, you cannot delete the root condition.

   • Delete a rule by clicking Delete Rule and then clicking Confirm.
4. Click Save.

To delete a Bot Manager configuration

1. Navigate to the Bot Manager page. How?From the main menu, navigate to More | Security | WAF Tier | Security Rule Manager | Bot Manager.

2. Click on the desired Bot Manager configuration.
3. Click Delete Bot Manager Config.
4. Type DELETE.
5. Click Delete.