Using an Agent

Now an RTA has been created its time to setup the Agent on a remote system or another server on the network. Obviously the RTA needs to be distributed to the location/system of choice. The RTA can be sent by email, posted for download, moved across a network (FTP) or even burned on to CD. The RoutePlotter Server is a fully functional Web Server and provides support for downloading an RTA via the browser if required without the use of a commercial web server such as Apache/IIS.

Once it has been copied to the system of choice, simply double-click the RTA ICON to execute as shown below:

When the Agent is started on the new machine the server that created the Agent will now register it as being online (shown below) as opposed to offline.

When the Agent is selected, as it is above, the Agent Summary box will give you 'last contact' details and how many routes it is running as shown below:

Installing the Agent as a service

The Agent can also be installed a service on the machine of choice. This can be done by clicking the 'Install agent as a service' link in the left hand side panel of the agent as shown below:

Once this has been done a confirmation dialog box will appear confirming that it has been installed as a service. Then, as with the main RoutePlotter, settings can be changed in Computer Management - see installing RoutePlotter as a service

Add a new route to an Agent

For security reasons adding or changing an RTA route cannot be done using the RTA agent directly and can only be done from the RoutePlotter Server panel.

To add a new route to an Agent click the ' View Agent ' link in the main Agents window as shown below

This will bring up the window below where the "Create a new route" link needs to be clicked, as shown in red below

Once 'Create a new Route' has been clicked the 'Trace Route' details dialog will appear as shown below

The options for this dialog box are explained below:

Description: the new route needs to be given a description that will appear at the beginning of each listing to identify one route from another.  It is advisable to use names that are brief as long names can make certain views look more complicated.

Address: the actual domain/IP for the trace route needs to be chosen. If the chosen route is an IP address then there is no need to click the 'Resolve' button; if the entry is a domain name, such as www.visualware.com, then the resolve button needs to be clicked in order to resolve the IP address for that domain.

Monitor Service Port: This provides a simple port testing facility. When selected the port option defines a specific port number that will be monitored as an additional test when tracing the route to the device. The selected port number defined is tested to make sure it is responding to Application requests.  For example using port 80 allows you to validate that the web server is responding on the IP device being tested.

Use UDP if ICMP traceroute fails: If ICMP fails to reach the device being monitored because of firewall issues RoutePlotter will try again using UDP as in some cases UDP may be allowed where ICMP is being blocked.

Interval: This option allows an interval to be set for when a new trace should be performed. The options range from 1 minute to 1 hour. When the interval expires RoutePlotter will refresh the trace route and update the database information accordingly.

There are some advanced options also available by clicking the Advanced button. The dialog box below will appear

This dialog allows the size of each ping packet to be set (bytes) and the ping timeout to be set (ms). The ping timeout value is the maximum time RoutePlotter will wait for a ping packet to return. Packets that exceed this time are reported as lost. The 'size' option allows you to set how many bytes will be sent in each ping packet. The default is 32. The minimum value is 4 bytes but setting higher values can be useful in gauging how applications are performing as this better mimics application data packets at the router.

Once the settings for the new route are complete click the Ok buttons until the Agents window is back in view. The new route will now be listed in the server routes list

The new route now appears in the 'Agent Routes' list as shown above. More options are now available in the left hand side column. These new options are edit route, view rules and delete route. By clicking 'View Rules' you can set a new rule for the Agent or edit old rules.

Setting a new rule for an Agent

First of all select the Agent route for the new rule as shown below. To get to this screen from the main RoutePlotter GUI click on the 'My Agents' link, then click the Agent in question and then click the 'View Agent' link. Then the screen below will appear

Select the Agent route wanted (red arrow above) then click the 'View Rules' link as shown above. This will show the dialog box below, where the 'Add New Rule' link needs to be clicked, this will bring up the window below.

When a new rule is added it will appear in the Private rules for... list to the right hand side of the dialog box. To create a new rule click the Add a new rule link to the left side of the dialog box. This will bring up the dialog box below

The new rule window consists of 4 steps, conditions, actions, exceptions and configure rule. When options are checked in the first 3 steps RoutePlotter adds each option to step 4. When the options for the new rule are chosen the rule now needs to be configured.

When a rule is added that contains a variable value item that needs to be set for the rule (such as a millisecond alarm value) it will be shown as a blue underline placeholder. To configure the rule just click the blue underlined placeholder. This will either rotate through some options such as low, normal and high for the priority setting or provide an edit box for a value to be entered such as an email address needed for an email alert rule.

If the without loose source routing through <an ip path> is selected the following dialog box will appear:

The 'Without loose source routing through' rule will be broken if the traceroute path doesn't include a set of IP's i.e. the route stops going through a server. Specify the IP's needed to route through by adding each IP to a different line in the text box. More than one IP on a line (separated by a space) means the route can go through any one of the IP's.

Note: The route table is only displayed if the route in question has been tested. In the case were an agent route is being selected and no agents are deployed the route table will not be displayed. IP's then need to be entered manually.