|
Jagger Homepage |
Jagger is a Java application monitoring tool using JMX technology to aggregate, archive and visualize monitoring data for larger computer clusters, giving developers and administrators both a succinct and comprehensive view into their systems, which normal JMX consoles cannot do due to information overflow.
Usage scenarios and system environment of Jagger.
The screenshot below is the Tomcat webapps panel of the console mode and shows how you can check on the health of your applications with one quick glance, unlike jconsole where it's buried in mostly irrelevant information.
More screenshots of other panels are available.
Started in demon mode, Jagger first reads the model definition describing the remote beans containing the raw monitoring attributes and possible aliases for computations on a scalar level, the target beans with the expressions to compute the aggregated attributes and a description of the cluster of JVMs that should be polled.
It then registers the target beans with the local JMX agent and thus exports the defined data model of aggregated values to any interested JMX client. The following diagram shows the resulting data flow for a typical scenario.
Data flow of aggregation.
It's also possible to periodically poll and dump a model definition to a set of CSV files for further batch processing.
First, get a source or binary (pre-built) release from the download pages. After installing (see also INSTALL.txt), you can directly start by calling the "jagger" command like this:
jagger -n lxjavatest9101:28113 -u monitor -w monitor
Check out the command line page to learn about available command line options. For more information on how to better adapt Jagger to your environment, see the configuration page.
