Sentinel3G User Guide
This page was last modified 05:26, 14 August 2006.From Documentation
| Revision as of 00:02, 2 May 2006 Daniels (Talk | contribs) (→Where to from here) ← Previous diff |
Current revision Daniels (Talk | contribs) (→The Console Window) |
||
| Line 10: | Line 10: | ||
| *the supplied knowledge bases (KBs) and how to add a new KB | *the supplied knowledge bases (KBs) and how to add a new KB | ||
| - | === Who should use this book === | + | === Who should use this guide === |
| There are two main audience groups: administrators set up and configure sentinel3G. They maintain details of sentries, agents, variables, etc.; operators monitor sentries and respond to events. Usually they monitor events via their console, but notification by other means such as e-mail may be used. Administrators are also sentinel3G users. | There are two main audience groups: administrators set up and configure sentinel3G. They maintain details of sentries, agents, variables, etc.; operators monitor sentries and respond to events. Usually they monitor events via their console, but notification by other means such as e-mail may be used. Administrators are also sentinel3G users. | ||
| Line 17: | Line 17: | ||
| === Installation === | === Installation === | ||
| - | Install sentinel3G on each Linux or UNIX server that you wish to monitor. You must run the installation script install.cos as root. Installation messages are logged to /tmp/FSinstall.log. <datetime> | + | Install sentinel3G on each Linux or UNIX server that you wish to monitor. Installation messages are logged to /tmp/FSinstall.log. <datetime> |
| The software distribution includes standard KBs (knowledge bases), which are installed automatically along with the software. Other KBs are available and can be installed separately. | The software distribution includes standard KBs (knowledge bases), which are installed automatically along with the software. Other KBs are available and can be installed separately. | ||
| Line 32: | Line 32: | ||
| You open a console window by running this command from a shell: | You open a console window by running this command from a shell: | ||
| - | cos sentinel | + | |
| + | <code>cos sentinel</code> | ||
| === Contacting Technical Support === | === Contacting Technical Support === | ||
| Line 79: | Line 80: | ||
| === The Console Window === | === The Console Window === | ||
| - | Operators and administrators monitor sentries through a console. Several people may each have one or more consoles, each displaying a different view of the events being monitored. You open a console window by running this command from a shell: | + | Operators and administrators monitor sentries through a console. Several people may each have one or more consoles, each displaying a different view of the events being monitored. You open a console window by running this command from a shell: <tt>cos sentinel</tt> |
| - | cos sentinel | + | |
| The console window looks like this: | The console window looks like this: | ||
| Line 118: | Line 118: | ||
| == Where to from here == | == Where to from here == | ||
| - | [[Sentinel3G Concepts ]] | + | [[sentinel3G Concepts]] |
| [[Monitoring Sentries From the Console]] | [[Monitoring Sentries From the Console]] | ||
Current revision
Contents |
Welcome to sentinel3G
Welcome to sentinel3G™ event monitoring and response software for monitoring networks of Linux, UNIX and Windows NT/2000/XP computers. Together, sentinel3G and this guide will help you to achieve agreed service levels, increase the return from IT assets, minimize down-time and maximize access to network services and resources.
About this Guide
This User Guide is the primary reference for finding out how to install, configure and use sentinel3G. The main sections describe:
- network monitoring concepts, and sentinel3G’s architecture and key components
- how to install and configure sentinel3G
- how to create and customize sentries
- how to use the console to monitor, diagnose, and respond to events.
- the supplied knowledge bases (KBs) and how to add a new KB
Who should use this guide
There are two main audience groups: administrators set up and configure sentinel3G. They maintain details of sentries, agents, variables, etc.; operators monitor sentries and respond to events. Usually they monitor events via their console, but notification by other means such as e-mail may be used. Administrators are also sentinel3G users.
Administrators have access both to a console and to the sentinel3G configuration options. Operators typically see only a console. Other interested parties include those with management or statutory responsibility for the procedures implemented in sentinel3G, including IT managers and auditors. They may or may not be sentinel3G users.
Installation
Install sentinel3G on each Linux or UNIX server that you wish to monitor. Installation messages are logged to /tmp/FSinstall.log. <datetime>
The software distribution includes standard KBs (knowledge bases), which are installed automatically along with the software. Other KBs are available and can be installed separately.
Users can access sentinel3G’s graphical user interface (GUI) from a Windows PC through the COSmanager PCgui for Windows. This software is also available from our downloads site. Install it on each Windows client.
Starting sentinel3G
The first time sentinel3G is started it will prompt for the name of the host running the Event Manager.
You can either select the current host or click Choose to choose from the entries in the hosts file.
Startup and Shutdown
sentinel3G’s data collection and monitoring engine includes the central Event Manager, and a host monitor running on each server. The installation procedure updates /etc/services and rc scripts so that these processes will be started automatically during each server’s boot process and stopped during each server’s shutdown sequence.
You open a console window by running this command from a shell:
cos sentinel
Contacting Technical Support
If you experience a problem, we suggest you first consult the user guide and release notes. If you have a support contract you will have been supplied with the support e-mail address which is the point of contact for questions about configuring and using sentinel3G. The support staff will be able to handle your question most effectively if you gather some key information first.
Hardware and software environment
- the version number of the product
- if you are running the licensed version, the license key and options used to license the product
- hardware type and configuration
- the operating system and version
Details of the problem or question
- what task were you performing or what process was running?
- what was the sequence of events before the problem occurred?
- what were the symptoms of the problem, including any error messages reported by the product or operating system?
- were you able to reproduce the problem?
Overview
sentinel3G is a smart, flexible monitoring technology for UNIX, Linux and Windows NT/2000/XP systems. It intelligently monitors and manages operating systems, SNMP-enabled devices, internet services, databases, and applications.
sentinel3G is easy to learn, easy to use, and easy to customize. The Sentinel3G graphical viewing console is accessible from any X Windows desktop or Microsoft Windows™ desktop.
This section gives an overview of the management philosophy underlying Sentinel3G, its graphical console and interface, and its features and day-to-day operations.
Management Overview
Sentinel3G is a tool to help system administrators and managers to measure, monitor and exceed agreed service levels. It is equally valuable whether run under the COSmanager environment, under some other system management framework, or as a standalone tool.
The COSmanager system management environment is aimed at organizations with small or large networks of UNIX and Linux servers, typically running RDBMS applications. COSmanager promotes system management based on policy and procedures rather than technology. Policy-based management ensures a consistent, reliable service, and COSmanager encourages data center staff to see themselves as service providers to end users.
Some aspects of this service are clearly visible to the end user. Is my application available? Is response time acceptable? Were batch transactions processed on time? Others are less obvious but no less important. These include efficient access to system resources, data integrity, fast recovery of files from backup, and many others.
What Sentinel3G Does
Sentinel3G provides a structured management environment in which to achieve service targets and satisfy service level agreements in modern business environments.
It allows centralized management of events and responses over a network.
Sentinel3G can automatically respond to events, or notify designated staff through a console or some external means if the problem requires a decision or action by a person.
Sentinel3G can be thought of as a data reduction filter. It collects masses of data from many hosts and presents only the key information necessary for the user to decide what to do (or decide that they don’t need to do anything).
Sentinel3G’s main goal is to free administrators and operators from routine, time-consuming tasks. If no system events need action, Sentinel3G is virtually invisible. If low-level events can be resolved safely, Sentinel3G attempts to do so.
If a problem requires human intervention, Sentinel3G notifies designated staff and presents any relevant information and options.
The Console Window
Operators and administrators monitor sentries through a console. Several people may each have one or more consoles, each displaying a different view of the events being monitored. You open a console window by running this command from a shell: cos sentinel
The console window looks like this:
Figure 1 — A console window
You can see at a glance the most serious alert in a folder; open the folder to find which sentries have alerts; display graphs and reports to help diagnose the problem; and run predefined actions to fix the problem and return the sentry to its normal state.
Which sentries and menu options are available to each user depends on their role. An operator can usually respond to events on selected sentries, but not change the sentries’ properties or reconfigure Sentinel3G itself. A Sentinel3G user with the Manager role would typically have full access to all sentries and menus.
Icons and overlays
Each icon represents a sentry, which is an instance of an object that is being monitored, or a folder, which contains a group of related sentries. A change in an icon’s color or appearance represents a change in the current state of the sentry. For example, if the available free space in a filesystem goes from 11% to 6%, and the sentry monitoring that filesystem changes from a warning state to alarm state, this would be shown by a change in the color of the sentry name and indicator icon from orange to red.
Monitoring sentries and responding to events
An operator can respond to an event by:
- viewing more information about the current state of the sentry in the form of a graph, report, or property sheet
- change the view (to a sub-folder or to Abnormal sentries view) to see if several alerts seem to be related to the same problem
- selecting one of the predefined actions to run
- acknowledging having seen the alert and having the sentry returned to normal state
Sentinel3G itself can respond to an event by:
- sending a notification message to one or more people via e-mail or pager.
- running an automatic response
Configuring Sentries
You can add new sentries to monitor some attributes of an object. A monitored object would typically be a database, device or process on a particular host. To do this, you need to define:
- an agent, which collects data on behalf of a sentry by running a command or by receiving data from a third-party application
- variables, which contain the latest data returned by the agent
- a sentry, which presents the current state of the object in a console
- the sentry’s states, which evaluate the variables to determine how severe the sentry’s condition is
You can also define additional information to help operators diagnose and respond to problems:
- graphs
- actions and reports
After defining some sentries you can run the host monitor in test mode to verify whether the agents are generating the correct data and that the state thresholds are appropriate.
Where to from here
Monitoring Sentries From the Console
Configuring Sentries and Agents