The Jeyzer Monitor Oneshot permits to execute Jeyzer Monitor in a one shot mode.
This is useful for the APM solutions like Nagios : the APM is requesting the state of the monitored process by calling the Jeyzer Monitor Oneshot.
Monitoring state is read from the process output (see Output section below).
The monitoring outcome is also published through HTML pages.
The monitor web publisher is therefore active and should have access to any HTTP server to publish those pages (by default the Tomcat instance shipped with Jeyzer).
Limitation : events are not persisted from one monitoring execution to another. Therefore, it is recommended to set a large time period between each execution call (like 15 minutes).
Command
To start the Jeyzer Monitor Oneshot execute the command jeyzer-monitor-oneshot.bat
or jeyzer-monitor-oneshot.sh
, located under the analyzer\bin
directory of your Jeyzer installation.
Prerequisites
These environment variables must be set to target the right profile :
JEYZER_TARGET_PROFILE
: the target master profileJEYZER_TARGET_NAME
: the application node name. Identifies the process instance.JEYZER_RECORD_DIRECTORY
: the Jeyzer recording directory to monitorJEYZER_RECORD_PERIOD
: the recording period
By default, the Jeyzer Monitor Oneshot targets the Jeyzer features demo application.
Installation
Jeyzer Installer : Select the Jeyzer Monitor component.
The Jeyzer Installer will guide you in the high level setup.
Check the quick_start.html
at the Jeyzer installation root : it contains useful paths.
Configuration
Jeyzer Monitor Oneshot configuration relies on :
- Monitoring profile, part of the master profile
- Monitoring rules
- License
Logging
Jeyzer Monitor Oneshot relies on the logback framework.
The nagios-logback.xml
drives the console and file logging.
It is located in the analyzer/config/log
directory of your Jeyzer installation.
Logging is disabled by default as the caller requires only on line of output.
Logging settings can be changed at runtime (no need to restart).
Log file is generated under the monitored directory recording which is by default :work/recordings/<application recording>/log
Outputs
This is basically adapted for the Nagios APM.
The execution output is one trace line – on the process stdout – which gives the status (no events/events/error) and refer a monitoring events doc URL if applicable.
The process exit code indicates the highest event category encountered : 0 for INFO, 1 for WARNING, 2 for CRITICAL and 3 for ERROR or UNKNOWN.
As a URL is exposed, HTML pages containing the monitoring events are also generated through the Jeyzer web publisher.