RTM configuration tool (rtminst)

Most of the configuration actions for the NAM Probe software are performed using the rtminst program. This program is installed as part of the installation package and can be accessed by the root user through the rtminst command.

You must have root user privileges to issue the rtminst command from the operating system prompt.

The main menu gives you access to specific NAM Probe configuration options. Each menu option has a number. To select an option, enter the corresponding number and press [Enter] . Enter [X] to exit to the main menu.
## 1 - Interface identification and network setup {#anchor_rtminst}

The network interface configuration is performed through the Interface identification and network setup on the main rtminst menu. You will choose one or more communication ports and identify other interfaces as the capture interfaces.

The NAM Probe machine is equipped with at least two network interfaces. Designate at least one NIC as a communication port that is used to log in remotely via SSH, send data to the report server, and communicate with the NAM Console software. Configure another NIC to collect traffic data (enter promiscuous mode).

The program checks the available drivers and interfaces and displays the current configuration information, as shown in the example below.

Note

Errors or missing configuration items are displayed as exclamation marks (!) and in the reference error messages displayed further down on the screen. The items changed since the last output refresh are marked with (*).

In the current configuration information, review the configuration changes.

Configuring drivers and interfaces is an iterative process of displaying the current configuration, making a change to it or making a change to the physical configuration by connecting or disconnecting a network cable, and then re-displaying the new configuration information. Therefore, after each configuration change you make, compare the expected result with the actual result shown in the updated screen information.

This process helps you associate the physical interfaces with particular entries in the rtminst table. The association is achieved by connecting a network cable to the interface and observing the interface being recognized by rtminst . The interface appears as Y in the Link det. column, and, if there is traffic on the interface, the traffic information appears in the Traffic column as In, Out or In+Out .

It is not necessary to disconnect all of the network cables before you can identify a particular interface. When you connect each successive cable to the interfaces, their entries are added to the table, allowing you to match the physical interfaces to the entries in the table. To refresh the table, press [Enter] after each physical change: after you plug in a cable, wait a few seconds then press [Enter] . It should take only a few seconds for the system to recognize a physical change made to one of the interfaces.

We recommend that traffic is present on each network cable used for the interface identification, as it can make it possible to identify the interfaces faster and more easily.

Note

Depending on traffic volume on a given interface, the table information might not always be updated to reflect the new state when you press [Enter] . If you have problems with a link or traffic not being detected, use the [I] # command to force rtminst to take extra time to perform additional tests on the interface and display the updated information.

Choose ports

Choose the port (or ports) to use for either capture, or communication. Leave at least one port available to configure as a capture port.

Note

Although it is possible to designate all of the NICs as communication ports, this configuration is not functional because no capture ports would be available on the device.

  • Identify the physical port.
    Identify the physical port by observing the changes in the table entries . Also, you can display the details about the interface and blink the interface LED (if equipped) using the[I] # option, where # is the number of the entry in the table of interfaces.
  • Select the port.
    Enter [M] # where # is the number of the entry in the table of interfaces, and then select S to configure the port to be a capture port, or U to set the port as unused.
  • Optional: Specify a port label for the specific capture port. Press press [Enter] to skip port labeling and accept blank port label.

Select full duplex speed or auto-negotiation. Enter [P]# where # is the number of the entry in the table and then select the appropriate option.

When a NAM Probe with 1 GB copper wire network cards is connected to a 100 Mbps tap or a switch mirror port and auto-negotiation fails, you can force the traffic sniffing devices to full duplex at 100 Mbps to avoid performance degradation.

I # - show details about interfaces

Use this option to view details of the selected interface. enter i and the corresponding interface number from the summary table.

T - change driver mode

In general, you should choose the Dynatrace custom network drivers. The current driver type is displayed on the rtminst output. To change the driver type, enter [T] and then select [C] for custom or [N] for native, as on the example below.

A - set all connected unused ports to capture mode

You can set all connected unused ports to capture mode:

  • Ensure that the link is detected for all of the relevant ports.
    Connect a network cable to the interface and observe the interface being recognized by rtminst . The interface will appear as Y in the Link det. column.
  • Select [A] to set all connected unused ports to capture mode.
    You are prompted to confirm your choice, after confirming, the ports listed as capture ports appear in the table.

S - save configuration and apply changes

Before exiting the setup, review all of the settings.

2 - AMD setup

1 - Data memory limit

The data memory limit is the maximum memory size that the NAM Probe process is allowed to use (in megabytes); it is pre-configured to an optimum value at NAM Probe software installation time but can be changed later if you add memory to the system.

RTMHS - The core NAM Probe monitoring process. Use the recommended value, which is 70% of the system RAM size.
page2trans - page2trans converter. Do not change.

The memory limits for v2page and page2trans converters are set to optimal values by default. Changing them is an advanced operation and should not be performed unless explicitly necessary.

2 - Driver parameters set

Driver parameters set, enables you to enter additional settings for a sniffing interface. If you need to pass additional parameters to the interface, select the Driver parameters set option, type [E] and, when prompted to enter the new value, press [Enter].

3 - RTM HS driver parameters

DO NOT CHANGE IF YOU DO NOT HAVE TO

The following settings are critical and have a high impact on NAM Probe monitoring performance and capabilities. They are designed to customize your driver settings and utilize advanced features of your hardware. The defaults selected for these settings are compatible with most common hardware and should not be changed.

1 - number of RX queues / workers

Default setting: [hardware dependent]
Set a number of CPU threads dedicated for traffic analysis. Possible values are from 1 to the number of CPUs or to 16 if you use the custom driver.

2 - RX queues buffer size

Default setting: [hardware dependent]
You can set the buffer size where the NAM Probe stores packets for analysis. Number of buffers corresponds to the number of threads defined in the Number of RX queues / workers setting.

3 - first affinity CPU

Default setting: 1
In the multicore architectures, set the first affinity CPU to be other than the default 1 and this will force the analyzing CPU threads to begin operating starting from a specified CPU.

4 - percentage of short packet buffers [ custom driver only ]

Default setting: 40%
This is the percentage of memory dedicated for short packets. Matching the appropriate amount of memory to the percent of short packets in the traffic allows you to use your memory more efficiently.

5 - software load balancing mode [ custom driver only ]

Default setting: false
You can set the method of load balancing between threads defined in the Number of RX queues / workers. Changing the setting, you define the load balancing to be performed by software instead of the hardware. This setting is relevant in deployments where your network interface card cannot perform this task.

6 - packet RX ring buffer size [ native driver only ]

Default setting: 200MB
You can set the amount of memory dedicated for packet exchange between system native driver and the NAM Probe software.

4 - Enabling non-secure data transfer over HTTP

Default setting: disabled
Enables or disables the non secure data transfer between the NAM Probe and the report server vie the HTTP (port 9091). To enable the HTTP communication, select Enabling non-secure data transfer over HTTP and then Enable non-secure data transfer.

5 - Enable automatic AMD setup

Default setting: disabled
This setting automatically adjusts values for:

  • Data memory limit
  • Driver parameters set
  • RTM HS driver parameters

These options will not be available in the rtminst menu while the automatic NAM Probe setup is enabled.