Solaris Network Statistics Plug-In Monitor
Purpose
Monitors network traffic on Solaris servers.
Description
This plug-in monitor retrieves network traffic information and triggers alerts based on byte counts for processed data and collisions, as well as straight error counts.
The monitor gathers these kernel statistics:
- rbytes
- obytes
- Collisions
- ierrors
- oerrors
Requirements
Monitoring station
up.time version – v4 or greater
Dependencies – Perl
Agent system
Platform – Solaris
Agent version – v4 or greater
Dependencies- Perl
Installation
Windows
- Install the downloaded plug-in monitor using loadpluginmonitor.exe
On the command line, run loadpluginmonitor.exe (found in the scripts directory), providing the absolute path to the downloaded plug-in monitor.
e.g.: loadpluginmonitor c:Program Filesuptime softwareuptime4scriptssolarisnetstats_windows.zip
or scriptsloadpluginmonitor c:solarisnetstats_windows.zip
Running loadpluginmonitor unarchives the plug-in monitor contents to the appropriate up.time /script and /xml directories (and if applicable, /agent and /lib directories), and will add the plug-in monitor to the up.time interface (specifically, the Advanced Monitors section of the Add Service Monitor page).
- Verify that the monitoring station Perl environment variable is set to c:perlbin
The plug-in monitor scripts assume that a Perl interpreter is located at, and the PERLDIR environment variable is set to, c:perl. If Perl is not found at this location, you must either install it there, or modify the PERLDIR path value in the solariszoneworkload.bat file so that it points to the actual Perl interpreter location.
Since all plug-in monitors are written with the assumption that a Perl interpreter is found at c:perl, it is recommended that you install Perl to that location.
UNIX
- Install the downloaded plug-in monitor using loadpluginmonitor
On the command line, run loadpluginmonitor (found in the /scripts directory), providing the absolute path to the downloaded plug-in monitor.
e.g.: loadpluginmonitor/usr/local/uptime4/scripts/solarisnetstats_unix.zip
or /scripts/loadpluginmonitor/user/local/solarsnetstats_unix.zip
Running loadpluginmonitor unarchives the plug-in monitor contents into the appropriate up.time /script and /xml directories (and if applicable, /agent and /lib directories), and will add the plug-in monitor to the up.time interface (specifically, the Advanced Monitors section of the Add Service Monitor page).
- Verify that the monitoring station Perl interpreter is located at usr/bin/perl
The plug-in monitor scripts assume that a Perl interpreter is located at usr/bin/perl, and will not work otherwise. If Perl is not found at this location, you must either install it there, or modify the path in the solariszoneworkload.pl hashbang so that it points to the actual Perl location.
Since all plug-in monitors are written with the assumption that a Perl interpreter is found at usr/bin/perl, it is recommended that you install Perl to that location.
Next Steps
This plug-in monitor includes an agent-side component that must be installed on the agent system from which data is to be collected.
To install the agent-side component on UNIX-based agent systems, perform the following steps:
- Locate the agent-side archive on the monitoring system.
When you unarchved the plug-in monitor using loadpluginmonitor, the agent-side archive was placed in the /scripts/Solaris Network Statistics/ directory. - Copy agent_solarisnetstats_solaris.tar.gz to the /opt/uptime-agent/scripts directory on the agent system.
- Unarchive the agent-side component into the /scripts/ directory.
- Ensure agent_solarisnetstats.sh can be executed by the up.time user (e.g., the script can be run by Nobody).
- Create a file named “.uptmpasswd” in the /opt/uptime-agent/bin directory.
This file should contain the agent password and absolute path to the agent-side script. - Ensure the times on the agent system and monitoring station are the same.
- Run the agent-side script:
# ./uptmagnt “rexec secret /opt/uptime-agent/scripts /agent_solarisnetstats.sh” where “secret” is the agent password.
Configuring the Monitor in up.time
The following configuration settings need to be set when creating a service instance based on this plug-in monitor:
- Script Name – the name and location (on the monitoring station) of the script that calls the agent-side script, which in turn determines what type of data is collected and sent back to the monitoring station
- Port – the port through which the up.time monitoring station communicates with the agent
- Agent Password – password required to access the up.time agent that is running on the monitored system
The following metrics are used to set alerting thresholds, and can be included in Report graphs:
- rbytes – the amount of data (bytes) read
- obytes – the amount of data (bytes) output
- Collisions – the number of collisions
- ierrors – the number of input errors
- oerrors – the number of output errors
Configuration Examples
This configuration is used on a Windows-based monitoring station. The plug-in monitor connects to the Solaris server and gathers the network statistics. If the number of input or output error thresholds are exceeded, the appropriate alert notifications and actions are carried out.
- Port – 9998
- Agent Password –
- Script Name – scriptsSolaris Network Statisticssolarisnetstats.bat
- ierrors – Warning: 200, Critical 300
- oerrors – Warning: 200, Critical 300