Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

Agent and agentless monitoring:  Many products on the market today claim to use agentless monitoring.  The only products that can truly make this claim are those products that sit out of band and monitor via network traffic.  True agentless monitoring has a drawback in that you can only gather information that crosses the network which does not tell you anything about the health of a system that delivers the service.  Zabbix is an agent based monitor that relies on the system itself or the monitoring server to detect problems, collect data and determine action based on this information.  In this case, Zabbix can use its own proprietary agent or , an SNMP agent or a program on the client or target system.

Types of Agents: Summary: A program that resides on the server gathering data on specific processes, resources and services and on queue or by prompt reports the summary back to a management station - distributed model.

Polling Agent:  A monitoring server is configured to  periodically poll an agent on the target server requesting on demand current information about a service, resource or process - can report status when the target is down but chatty.

Trapping Agent:  A process running on the target server sends information to a monitoring server when an event is triggered - low overhead but can't report if the target is impaired or down.

Zabbix Agent Installations:

  1. Windows installations are the most simple to perform and are done using an MSI that is executed on the target system. \\rnas.rice.edu\ZabbixAgents
  2. Linux and Unix installation are only slightly more complex and can be accomplished by following the steps in Zabbix Agent .
  3. Firmware based networked devices such as printers and network gear can not run a proprietary agent and generally come with SNMP installed as part of the firmware. (RO community string = public, monitor.rice.edu = SNMP management station - please take not of the version of SNMP that your equipment support  V1, V2C, or V3.
  4. After installing and starting the agent on an operating system, the Zabbix service can be configured to automatically detect the new server, add the host in to the database and associate it within the appropriate host group.

NOTE:  Host Groups are important for access controls within the Zabbix permissions.  If your host is installed in the wrong host group, you will not be able to manage it until the Host Group entry for that Host has been modified.

Using an SNMP Agent:  While the Zabbix agent is useful and affective in adding a target to the server database automatically and placing it in the appropriate group, it is not required in order to monitor a system or service which can be done manually or via xml import for standardized gear such as network components, printers, storage systems or other closed operating systems or firmware.  Manually adding target items in Zabbix is done by configuring a new HOST entry.  The minimum required information for adding a new host is

...

  1. Non Authenticated: The web service will display information, graphs and some reports without authentication but no portal  information can be displayed since portals are user defined. 
  2. Authenticated:  This is the most useful way to access the service.  The layout for the services is hierarchically tab based.
    1. Click the login item in the far upper right of the screen to authenticate
    2. The display will be lite until you choose from the Tab hierarchy.
      1. Monitoring:  Provides display only pages with some display configuration for the following.  Most are Host Group organized and can be column sorted.
        1. Dashboard:  your personal portal dashboard that can be configured to display favorite graphs, screens and maps - Screens will be the most useful.
        2. Overview: A good way to do a quick status check per Host Group by comparing triggers to hosts.
        3. Web:  Graphic display and status of web based data such as download speed for a specific page etc..
        4. Latest Data:  The most recently polled data including graphs and trigger information that is Host Group searchable. This included items details.
        5. Triggers:  The most recent triggers searchable by Host Group.
        6. Event: A summary of the latest events of the current date by discovery (poll) or trigger.
        7. Graph: A searchable index of any graphs.  Graphs are created by stewards.
        8. Screen: A searchable index of grouped graphs is a screen.  This page provides an index of screens.
        9. Map: Not currently used, but can be configured to show network or other hierarchy in a graphical format.
        10. Discovery: A view of discovery rules that define how newly identified Zabbix Agents are configured into Zabbix automatically.
        11. IT Services: A view of SLA rules that allows automatically defining service triggers to uptime reports.
      2. Inventory
        1. Hosts:  A display of all servers and devices being monitored.  Hosts are the  basic unit of configuration in Zabbix
      3. Reports:
        1. Status of Zabbix:  The health of the monitoring system itself.
        2. Availability Report:  What services and systems are available - canned report by host or trigger template.
        3. Most busy triggers:  Canned report on which are the most active triggers.
        4. Bar Reports:  Canned bar graph of status of systems.
      4. Configuration:  (where your will do most of your work)
        1. Host Group:  Mapping of hosts to host groups.
        2. Host:  Individual machines on which services are monitored and where applications, items, graphs and triggers are configured.
        3. Maintenance: Where maintenance windows are defined for SLAs.
        4. Web: Specific Web monitoring configurations.
        5. Actions: Configuration for what action to take when a trigger is pulled.
        6. Screens: Configuration of graphical screens.
        7. Maps: Configuring graphical maps.
        8. IT Services: Configuration of SLA critieria.
        9. Discovery: Configuration of auto discovery - what to do when a host reports to the server and it is unknown.
        10. Export/Import: A location to export and import XML templates for host configurations including triggers, graphs, etc..