Variables in correlators

If tracking values in event fields, active lists, or dictionaries is not enough to cover some specific security scenarios, you can use global and local variables. You can use them to take various actions on the values received by the correlators by implementing complex logic for threat detection. Variables can be declared in the correlator (global variables) or in the correlation rule (local variables) by assigning a function to them, then querying them from correlation rules as if they were ordinary event fields and receiving the triggered function result in response.

Usage scope of variables:

• When searching for grouping or unique field values in correlation rules.
• In the correlation rule selectors, in the filters of the conditions under which the correlation rule should be triggered.
• When enriching correlation events. Select Event as the source type.
• When filling active lists with values.

Variables can be queried the same way as event fields by preceding their names with the $ character.

Properties of variables

Local and global variables

The properties of global variables differ from the properties of local variables.

Global variables:

• Global variables are declared at the correlator level and are applied only within the scope of this correlator.
• The global variables of the correlator can be queried from all correlation rules that are specified in it.
• In standard correlation rules, the same global variable can take different values in each selector.
• It is not possible to transfer global variables between different correlators.

Local variables:

• Local variables are declared at the correlation rule level and are applied only within the limits of this rule.
• In standard correlation rules, the scope of a local variable consists of only the selector in which the variable was declared.
• Local variables can be declared in any type of correlation rule.
• Local variables cannot be transferred between rules or selectors.
• A local variable cannot be used as a global variable.

Variables used in various types of correlation rules

• In operational correlation rules, on the Actions tab, you can specify all variables available or declared in this rule.
• In standard correlation rules, on the Actions tab, you can provide only those variables specified in these rules on the General tab, in the Identical fields field.
• In simple correlation rules, on the Actions tab, you can provide only those variables specified in these rules on the General tab, in the Inherited Fields field.

Requirements for variables

When adding a variable function, you must first specify the name of the function, and then list its parameters in parentheses. Basic mathematical operations (addition, subtraction, multiplication, division) are an exception to this requirement. When these operations are used, parentheses are used to designate the severity of the operations.

Requirements for function names:

• Must be unique within the correlator.
• Must contain from 1 to 128 Unicode characters.
• Must not begin with the character $.
• Must be written in camelCase or CamelCase.

Special considerations when specifying functions of variables:

• The sequence of parameters is important.
• Parameters are separated by a comma: ,.
• String parameters are passed in single quotes: '.
• Event field names and variables are specified without quotation marks.
• When querying a variable as a parameter, add the $ character before its name.
• You do not need to add a space between parameters.
• In all functions in which a variable can be used as parameters, nested functions can be created.

Functions of variables

Operations with active lists and dictionaries

"active_list" function

Gets information from the active list regarding the value in the specified column.

You must specify the parameters in the following sequence:

1. Name of the active list
2. Name of the active list column
3. Active list record key

   The name of one or more event fields is used as the record key of the active list.

   Usage example	Result
   active_list('exampleActiveList', 'score', SourceAddress,SourceUserName)	Gets data from exampleActiveList from the SourceAddress,SourceUserName record in the score column.

"table_dict" function

Gets information about the value in the specified column of a dictionary of the table type.

You must specify the parameters in the following sequence:

1. Dictionary name
2. Dictionary column name
3. Dictionary row key

   Usage example	Result
   table_dict('exampleTableDict', 'office', SourceUserName)	Gets data from the exampleTableDict dictionary from the row with the SourceUserName key in the office column.

"dict" function

Gets information about the value in the specified column of a dictionary of the dictionary type.

You must specify the parameters in the following sequence:

1. Dictionary name
2. Dictionary row key

   Usage example	Result
   dict('exampleDictionary', SourceAddress)	Gets data from exampleDictionary from the row with the SourceAddress key.

Operation with rows

"len" function

Returns the number of characters in a string.

A string can be passed as a string, field name or variable.

"to_lower" function

Converts characters in a string to lowercase.

A string can be passed as a string, field name or variable.

"to_upper" function

Converts characters in a string to uppercase. A string can be passed as a string, field name or variable.

"append" function

Adds characters to the end of a string.

You must specify the parameters in the following sequence:

1. Original string.
2. Added string.

Strings can be passed as a string, field name or variable.

"prepend" function

Adds characters to the beginning of a string.

You must specify the parameters in the following sequence:

1. Original string.
2. Added string.

Strings can be passed as a string, field name or variable.

"substring" function

Returns a substring from a string. 

You must specify the parameters in the following sequence:

1. Original string.
2. Substring start position (natural number or 0).
3. (Optional) substring end position.

Strings can be passed as a string, field name or variable. If the position number is greater than the original data string length, an empty string is returned.

"tr" function

Deletes the specified characters from the beginning and end of a string.

You must specify the parameters in the following sequence:

1. Original string.
2. (Optional) string that should be removed from the beginning and end of the original string.

Strings can be passed as a string, field name or variable. If you do not specify a string to be deleted, spaces will be removed from the beginning and end of the original string.

"replace" function

Replaces all occurrences of character sequence A in a string with character sequence B.

You must specify the parameters in the following sequence:

1. Original string.
2. Search string: sequence of characters to be replaced.
3. replacement string: sequence of characters to replace the search string.

Strings can be passed as a string, field name or variable.

"regexp_replace" function

Replaces a sequence of characters that match a regular expression with a sequence of characters and regular expression capturing groups.

You must specify the parameters in the following sequence:

1. Original string.
2. Search string: regular expression.
3. replacement string: sequence of characters to replace the search string, and IDs of the regular expression capturing groups.

Strings can be passed as a string, field name or variable. Unnamed capturing groups can be used.

"regexp_capture" function

Gets the result matching the regular expression condition from the original string.

You must specify the parameters in the following sequence:

1. Original string.
2. Search string: regular expression.

Strings can be passed as a string, field name or variable. Unnamed capturing groups can be used.

Operations with timestamps

now function

Gets a timestamp in epoch format. Runs with no arguments.

"extract_from_timestamp" function

Gets atomic time representations (year, month, day, hour, minute, second, day of the week) from fields and variables with time in the epoch format.

The parameters must be specified in the following sequence:

1. Event field of the timestamp type, or variable.
2. Notation of the atomic time representation. This parameter is case sensitive.

   Possible variants of atomic time notation:

   • y refers to the year in number format.
   • M refers to the month in number notation.
   • d refers to the number of the month.
   • wd refers to the day of the week: Monday, Tuesday, Wednesday, Thursday, Friday, Saturday, Sunday.
   • h refers to the hour in 24-hour format.
   • m refers to the minutes.
   • s refers to the seconds.
3. (optional) Time zone notation. If this parameter is not specified, the time is calculated in UTC format.

   Usage examples

   extract_from_timestamp(Timestamp, 'wd')

   extract_from_timestamp(Timestamp, 'h')

   extract_from_timestamp($otherVariable, 'h')

   extract_from_timestamp(Timestamp, 'h', 'Europe/Moscow')

"parse_timestamp" function

Converts the time from RFC3339 format (for example, "2022-05-24 00:00:00", "2022-05-24 00:00:00+0300) to epoch format.

"format_timestamp" function

Converts the time from epoch format to RFC3339 format.

The parameters must be specified in the following sequence:

1. Event field of the timestamp type, or variable.
2. Time format notation: RFC3339.
3. (optional) Time zone notation. If this parameter is not specified, the time is calculated in UTC format.

   Usage examples

   format_timestamp(Timestamp, 'RFC3339')

   format_timestamp($otherVariable, 'RFC3339')

   format_timestamp(Timestamp, 'RFC3339', 'Europe/Moscow')

"truncate_timestamp" function

Rounds the time in epoch format. After rounding, the time is returned in epoch format. Time is rounded down.

The parameters must be specified in the following sequence:

1. Event field of the timestamp type, or variable.
2. Rounding parameter:
   • 1s rounds to the nearest second.
   • 1m rounds to the nearest minute.
   • 1h rounds to the nearest hour.
   • 24h rounds to the nearest day.
3. (optional) Time zone notation. If this parameter is not specified, the time is calculated in UTC format.

   Usage examples	Examples of rounded values	Usage result
   truncate_timestamp(Timestamp, '1m')	1654631774175 (7 June 2022, 19:56:14.175)	1654631760000 (7 June 2022, 19:56:00)
   truncate_timestamp($otherVariable, '1h')	1654631774175 (7 June 2022, 19:56:14.175)	1654628400000 (7 June 2022, 19:00:00)
   truncate_timestamp(Timestamp, '24h', 'Europe/Moscow')	1654631774175 (7 June 2022, 19:56:14.175)	1654560000000 (7 June 2022, 0:00:00)

"time_diff" function

Gets the time interval between two timestamps in epoch format.

The parameters must be specified in the following sequence:

1. Interval end time. Event field of the timestamp type, or variable.
2. Interval start time. Event field of the timestamp type, or variable.
3. Time interval notation:
   • ms refers to milliseconds.
   • s refers to seconds.
   • m refers to minutes.
   • h refers to hours.
   • d refers to days.

   Usage examples

   time_diff(EndTime, StartTime, 's')

   time_diff($otherVariable, Timestamp, 'h')

   time_diff(Timestamp, DeviceReceiptTime, 'd')

Mathematical operations

These are comprised of basic mathematical operations and functions.

Basic mathematical operations

Operations:

• Addition
• Subtraction
• Multiplication
• Division
• Modulo division

Parentheses determine the sequence of actions

Available arguments:

• Numeric event fields
• Numeric variables
• Real numbers

  When modulo dividing, only natural numbers can be used as arguments.

Usage constraints:

• Division by zero returns zero.
• Mathematical operations between numbers and strings return zero.
• Integers resulting from operations are returned without a dot.

  Usage examples (Type=3; otherVariable=2; Message=text)	Usage result
  Type + 1	4
  $otherVariable - Type	-1
  2 * 2.5	5
  2 / 0	0
  Type * Message	0
  (Type + 2) * 2	10
  Type % $otherVariable	1

"round" function

Rounds numbers. 

Available arguments:

• Numeric event fields
• Numeric variables
• Numeric constants

  Usage examples (DeviceCustomFloatingPoint1=7.75; DeviceCustomFloatingPoint2=7.5 otherVariable=7.2)	Usage result
  round(DeviceCustomFloatingPoint1)	8
  round(DeviceCustomFloatingPoint2)	8
  round($otherVariable)	7

"ceil" function

Rounds up numbers.

Available arguments:

• Numeric event fields
• Numeric variables
• Numeric constants

  Usage examples (DeviceCustomFloatingPoint1=7.15; otherVariable=8.2)	Usage result
  ceil(DeviceCustomFloatingPoint1)	8
  ceil($otherVariable)	9

"floor" function

Rounds down numbers.

Available arguments:

• Numeric event fields
• Numeric variables
• Numeric constants

  Usage examples (DeviceCustomFloatingPoint1=7.15; otherVariable=8.2)	Usage result
  floor(DeviceCustomFloatingPoint1)	7
  floor($otherVariable)	8

"abs" function

Gets the modulus of a number.

Available arguments:

• Numeric event fields
• Numeric variables
• Numeric constants

  Usage examples (DeviceCustomNumber1=-7; otherVariable=-2)	Usage result
  abs(DeviceCustomFloatingPoint1)	7
  abs($otherVariable)	2

"pow" function

Exponentiates a number.

The parameters must be specified in the following sequence:

1. Base. Real numbers
2. Power. Natural numbers.

Available arguments:

• Numeric event fields
• Numeric variables
• Numeric constants

  Usage examples

  pow(DeviceCustomNumber1, DeviceCustomNumber2)

  pow($otherVariable, DeviceCustomNumber1)

Declaring variables

To declare variables, they must be added to a correlator or correlation rule.

To add a global variable to an existing correlator:

1. In the KUMA web interface, under Resources → Correlators, select the resource set of the relevant correlator.

   The Correlator Installation Wizard opens.

2. Select the Global variables step of the Installation Wizard.
3. click the Add variable button and specify the following parameters:
   • In the Variable window, enter the name of the variable.

     Variable naming requirements

     • Must be unique within the correlator.
     • Must contain from 1 to 128 Unicode characters.
     • Must not begin with the character $.
     • Must be written in camelCase or CamelCase.
   • In the Value window, enter the variable function.

     Description of variable functions.

   Multiple variables can be added. Added variables can be edited or deleted by using the icon.

4. Select the Setup validation step of the Installation Wizard and click Save.

A global variable is added to the correlator. It can be queried like an event field by inserting the $ character in front of the variable name. The variable will be used for correlation after restarting the correlator service.

To add a local variable to an existing correlation rule:

1. In the KUMA web interface, under Resources → Correlation rules, select the resource of the relevant correlation rule.

   The correlation rule settings window opens. The parameters of a correlation rule can also be opened from the correlator to which it was added by proceeding to the Correlation step of the Installation Wizard.

2. Open the Selectors tab.
3. In the selector, open the Local variables tab, click the Add variable button and specify the following parameters:
   • In the Variable window, enter the name of the variable.

     Variable naming requirements

     • Must be unique within the correlator.
     • Must contain from 1 to 128 Unicode characters.
     • Must not begin with the character $.
     • Must be written in camelCase or CamelCase.
   • In the Value window, enter the variable function.

     Description of variable functions.

   Multiple variables can be added. Added variables can be edited or deleted by using the icon.

   For standard correlation rules, repeat this step for each selector in which you want to declare variables.

4. Click Save.

The local variable is added to the correlation rule. It can be queried like an event field by inserting the $ character in front of the variable name. The variable will be used for correlation after restarting the correlator service.

Added variables can be edited or deleted. If the correlation rule queries an undeclared variable (for example, if its name has been changed), an empty string is returned.

If you change the name of a variable, you will need to manually change the name of this variable in all correlation rules where you have used it.