- Access your RT controller's SSH.
- Stop any ongoing tracing and clear the CPUs' buffers by executing the following commands
echo 0 > /sys/kernel/debug/tracing/tracing_on
echo 0 > /sys/kernel/debug/tracing/trace

- Adjust the size of the CPUs buffers for capturing traces
echo <new size in kB> | tee /sys/kernel/debug/tracing/buffer_size_kb
Where <new size in kB> represents the new size in kilobytes (kB) that is going to be used for all CPU cores. When setting this value, please consider the following:
-
-
- The size must be a multiple of 4 kB.
- The maximum size depends on the available memory on your RT controller. Be cautious when configuring large buffer sizes, as RAM is limited and shared with the rest of the system.
- The size configured through the command above is per CPU (core). For example, if your controller has 4 cores and you set 32,768 kB, it actually reserves 4 x 32 MB ≈ 128 MB in total.
- This configuration does not persist across reboots.

4. Verify that the new size has been set correctly by running the command shown below. The console output should display a value close to the one you configured.
cat /sys/kernel/debug/tracing/buffer_size_kb

5. Capture the required traces from your system, as explained in the Tracing on NI Linux Real-Time guide.