Cricket is a versatile and powerful system designed for tracking time-series data trends with high accuracy and flexibility.
The configuration files utilized by Cricket are called config trees, expressing everything the program needs to know about getting data for the system, the types of data it needs, and the targets to collect data from. The config trees are designed to eliminate redundancy, simplify management, and curb mistakes resulting from copy-and-paste errors. Cricket is developed in Perl and falls under the GNU General Public License, and is developed on Solaris machines under Apache. Although it is known to function on HP-UX, Linux, and various forms of BSD, as well as other operating systems, some users successfully operate Cricket on Windows NT and/or Windows 2000 but are yet to document it.
In the latest release of Cricket, the following updates were made: case dependency errors with new styles views were eliminated; a "label" tag was added to the "view" dictionary, reducing the need to use spaces in view names and permitting the Cricket admin to alias different targetTypes to the same name in the HTML menu; systemPerfConf.pl was reverted to using our SNMP_Session based SNMP interface, and interfaces detection was added. The distributed sample-config/Defaults removed width-hint and height-hint commentary preventing crushed graphics; NaN handling was made uniform throughout the code, with the use of the Common::Util's isNaN function to check for NaN; all text files now have LF line endings, while version 1.0.4 had CRLF line endings in some documents. The threshold monitoring team identified and resolved a small bug, while the monitoring subsystem has undergone extensive error reporting enhancements based on a patch by Andrew Clark. s390x-linux is now supported thanks to Shane Stixrud's Patch #877924.
Version 1.0.5: N/A