How to use System Variables in Threshold/Analyst alert message to make it more informative?

[David_Sun]

Prognosis provides a few built in System Variables that can be used in Threshold or Analyst alert messages to make it more informative about the detected event.

Threshold

^srcnode@  - Includes the node name where the data is being collected.
^currtime@  - Current timestamp. 
^ret_code@  - Return code indicating the success or failure of any EXEC or START command.
^ret_text@  - Any text that is returned after EXEC or START command.
^cur_prob@ - The automatically assigned problem number for this instance of the alert.

Note: the @ symbol suppresses trailing spaces.

Example of alert text containing variables:

"^srcnode@ CPU @NTCPU.CPUNO@ is @NTCPU.BUSY@% busy"

This would output:
 
"\MyNodeName CPU 0 is 80% busy"
 

Analyst

^srcnode  - Node where the data for this rule instance is being collected.
^currtime - Current timestamp.  
^diskint  - Current disk collector interval.
^ret_code  - Return code indicating success or failure of an EXEC or START command.  
^ret_text  - Text that is returned after EXEC or START command.
^cur_prob  - The auto assigned problem number for this instance of the rule.
 
Note: the @ symbol suppresses trailing spaces.

[ChristopherS]

In Analysts there are also ^Var0-30 that can be utilized with a MATCHES REGEX statement to parse output command output and place it inside of analyst variables. 

As of Prognosis 11.5p2 the SET command automatically performs a CAST function from Numeric or Text based on the analyst variable definition the ^var value is copied into for actual use (you technically can use it as a ^Var as well but it assumes text unless cast into a numeric)