Table of Contents
Alarms also make use of Targets but instead of using the value obtained from the Target to draw a graph, the value is compared to specified values, and depending on what type of comparison is done, the alarm will be active or not. An Alarm is usually in one of three states:
Table 9.1. Alarm States
|OK||Indicates that the test has passed and the service is operating as expected|
|Warning||There is a problem but the service may still be functioning|
|Critical||The test has failed.|
Any change in the state of an Alarm Target can trigger a 'Failure Action'. These are scripts that can perform a simple operation like sending an email to an administrator informing them of the problem, to an a more complex operation like actually trying to fix the problem.
Also See: Chapter 7, Targets
There are two ways to create a new alarm. The first is using the New Alarm Wizard, and the second is to copy and paste an existing alarm (from the Alarm List Window or the Properties Window). Once pasted the setting for the alarm can be edited to what is required. The other alternative is to use the New Alarm Wizard as described below.
To launch the New Alarm Wizard you could:
In broad terms the wizard steps are as follows:
You may have noticed the similarity between these steps and the New Graph Wizard's steps. They are very similar. The first difference is that instead of requiring an entry for graph info, this wizard requires an entry for Alarm info. The second and last difference is that an alarm consists of only one target, so here the option to add additional targets does not exist. The name of the target is, in this case the name of the alarm.
The option to choose one of the four aquirer types is offered:
If you select the Remote-Script then you will need to enter or select the computer that will run the script. If NetProbe-Remote is already setup on the required computer the list will show the computers IP Address. If you intend to setup the NetProbe-Remote.pl or NetProbe-Remote-Service later, then you will have to enter the computer name/IP manually.
The option to 'Do a Basic Test on Located scripts' is also offered in the case of the Script Aquirer. The test will often help determine if the script has a chance of working by testing for syntax errors and missing modules.
The information required by an aquirer is specific to the type of aquirer.
Settings for the PerfMon take two steps. The first is to select the host that will be tested. This could be a new host, requiring a new aquirer, or it could be a host that already has a PerfMon aquirer setup. In this second option you are adding a Slave Target to an existing aquirer. Slave Targets cannot have their Refresh Rate or Timeout set to anything other than that selected in the Primary Target. If these values are not suitable then use the 'Create New Aquirer' option. From a performance perspective re-using existing aquirers by adding Secondary Targets is most advantageous.
The next step (after the 'Next' is clicked), will show a list of Performance Objects. It may take a few seconds for this list to be obtained. Selecting the Object will result in the Counter and Instance blocks being filled in with the relevent info. Select the Counter, and if present an Instance, which is required to complete the PerfMon Target settings.
As with the PerfMon aquirer the SNMP also requires two steps. The first is to select whether to Create a New Aquirer or to use an Existing Aquirer. Once again to re-use an existing aquirer means you will be creating a Secondary Target and your Refresh Rate and Timeout will be dictated by the Primary Target in the aquirer. The 'New Host' option means a new Aquirer will be created. Before you can continue, in the case of a Create New Aquirer, you must enter the new host's community and IP or FQDN.
Clicking on the 'Next' button will take you to the next step in the setup of the aquirer. This is the selection of a MIB. All MIB's are displayed and you will be required to select the desired one. Only MIB's of the following type can be used:
Also See: the section called “SNMP Browser”
The aquirer setting for Script and Script-Remote are almost the same. The only difference is where the script is executed. In the case of Script, it is executed on the computer running Net-Probe. In the case of Script-Remote the script is executed on the host selected on the 'Select Type' page, under the 'Remote Host' option.
Your first choice in setting up these aquirers is to select if you wish to make use of an existing aquirer or if you want to create a new aquirer. As with all other types, using an existing aquirer means you cannot specify the Refresh Rate and the Timeout. This option does however have performance advantages. Selecting a New Script means a new aquirer will be created and you will be required to specify the Refresh Rate and Timeout. If a new script is chosen then you can simply click on the check box of one of the Located Scripts, or you can browse for a script anywhere on your file system. The located scripts are in the Net-Probe installed directory under 'ScriptsTargets' directory. Under this directory a number of other directories are shown, each the name of a Script Engine. All located scripts are placed within these directories. The Wizard uses the name of the Parent Directory to determine which script language to use. If you browse and select a script you will be required to select the script engine required or enter the name of a different script engine (you will only be permitted to enter, as opposed to selecting from the list, the name of the engine if the script is of type Script-Remote).
The Locate Scripts will offer some additional information about the script. If the 'Perform Basic Test on Script' was previously selected, then some of the scripts may have a red background and will not offer a checkbox. These scripts have failed the basic test. The last column in the list will offer some information as to why the test failed. The test is recommended, in the case of Script Aquirers. There are however additional reasons why a script may not be selectable, i.e. the script is only suitable for graphing.
Clicking on the Next option takes you to the Script Arguments section of the wizard. Here you are required to enter information that will be fed into the script. You will also be required to select the Return Arg. This specifies which of the return values would be used to create your graph. You will also be asked to select or enter the host which will be passed to the script.
If one of the 'Located Scripts' was selected then the wizard will make things a lot easier for you. Possible return arguments will be named and limited to what is applicable. Only required In-Args will be enabled. They will also be labeled with a descriptive name indicating what is required. This makes it very easy for anyone to make use of the pre-written scripts even if they know nothing what so ever about scripting.
Here you will be required to enter the 'Comparison Operator', the 'Comparison Values' (two: one for Warning and one for Critical), and the option to select an Alarm Action Script/s. The comparison Operator defines how the given value is compared to the target's return value. If for example the operator is = then the critical and warning values are compared as to whether that are exactly equal to the returned value of the target. If the target meets the requirements set for the Warning or Critical alarms then the Action Script is run (if it is enabled).
Additionally you may have the option to select a Workspace depending on how the wizard was started. The final option would be to specify if the alarm 'Is the Host Test'. This option will only be available if there is currently not an existing host test set for host on the Workspace.
|Copyright (c) Warren Flemmer 2006||www.net-probe.com|