NeoLoad provides two main and distinct applications, the NeoLoad Controller and the Load Generator. The latter, the Load Generator does the actual job of running a scenario by sending the requests it defines. The Controller provides NeoLoad main GUI, letting you design and launch scenarios and analyze their results. The Controller also manages the Load Generators, controlling them and delegating the running of the scenario to them. The Controller comes with its own built-in Load Generator.
The Load Generator actually comes with an additional executable called the Agent. The Agent is responsible for launching the Load Generator executable and for informing the Controller that a Load Generator is present on the machine.
General
The installation process closely resembles that of the Controller, the main difference being that you do not need to enter a license key. The NeoLoad license key operates on a per-Controller basis. You can have one Controller and as many Load Generators as you need. Refer to the previous diagram for a better understanding of NeoLoad deployment and license key requirements.
Also, keep in mind that a Load Generator is never launched manually. As previously shown, it is always started through the Agent.
During the Load Generator installation process, the wizard asks you whether the Agent should be run as a service and if it should be run on start up. NeoLoad default setting is to both install the Agent as a service and run it on start up. This default setting is usually the most appropriate.
If you decide otherwise, you will need to start the Agent. For more information, see Manage the Load Generator Agent<in_pdf_install>.
Firewalls
Since the Controller and Agent communicate with each other, any intervening firewall must be correctly configured. Three ports need to be opened: one for the Agent, one for the Load Generator, and one for the Controller. One UDP port must be defined for the Agent to allow the Controller to discover the Load Generators on the network. You may change all these default settings by editing the Controller or Agent configurations. For more information, see Use a firewall between the Controller and a Load Generator<in_pdf_install>.
To use NeoLoad with a firewall, it may prove necessary to open some ports on the network interface of the Controller, Load Generator, and Monitoring Agent. For more information, see Use a firewall between the Controller and a Load Generator and Use a firewall between the Controller and a Monitoring Agent.
Default port |
Protocol |
Configuration file |
7100 or configured as explained in Add Load Generators |
TCP |
Load Generator:
Controller:
|
Default port |
Protocol |
443 or configured as explained in Add Load Generators |
TCP |
Port |
Protocol |
Configuration file |
1359 |
UDP |
Load Generator:
Controller:
|
Port |
Protocol |
Configuration file |
7200 |
TCP |
Monitoring Agent:
Controller:
|
Port |
Protocol |
Configuration file |
1358 |
UDP |
Monitoring Agent:
Controller:
|
Port |
Protocol |
Configuration file |
4569 |
TCP |
Controller:
|
Information: To use the automatic discovery function of the Load Generators and Monitoring Agents, it is necessary to open the suitable ports on the hosting machines and for the Controller. If the ports are not opened, the Load Generators and the Monitoring Agents must be added manually, as described in Manage the Load Generator Agent and Manage the Monitoring Agent.
If you have installed several Load Generators and their associated Agents, you may launch a scenario using these Load Generators. In the Scenario panel in the Runtime section, click on the Discover button (two arrows symbol) to display all the Load Generators installed and currently available. In the example, NeoLoad displays the built-in Generator (localhost), an available Load Generator discovered on a machine whose address is DEXTER
, and a Generator named loki
that was discovered earlier and which is no longer available.
If you select a Load Generator and then click on the Advanced button, NeoLoad displays the Advanced Host Configuration dialog box. In the Network tab you may select the specific network interfaces, (i.e. the interfaces to the server hosting the application to be tested) the Load Generator should use to replay a scenario. Likewise, you can specify which IP address the Load Generator should show to the tested application when sending HTTP requests.
The Load Balancing tab lets you define how much load to generate on that particular Load Generator. If all the Load Generators have a load factor of 1, they will all generate the same load. Conversely, if one Load Generator has a load factor of 2, while all the others have a load factor of 1, then first Load Generator will generate twice as heavy a load as the others.
Example of load balancing according to load factor. Given that there are two Generators: Generator 1 and Generator 2:
Consequently, Generator 1 will generate 1/(1+3) = 1/4 = 25% of the load and Generator 2 will generate 3/(1+3) = 3/4 = 75% of the load.
For more information about installing Load Generators, see Zones and Load Generator hosts.