Advanced configuration
editAdvanced configuration
editAfter completing the steps in the Get started documentation, you may need to continue to more advanced configuration for your deployment. See the following sections for more information:
- Tag data for querying: Tag data collected by the Universal Profiling Agent into multiple logical groups so they can be queried in Kibana.
- Add symbols for native frames: Push symbols to your cluster so you can see function names and line numbers in traces of applications written in programming languages that compile to native code (C, C++, Rust, Go, etc.).
- Use a proxy: Set up an HTTP proxy if your infrastructure Universal Profiling Agent installation needs one to reach Elastic Cloud.
- Override kernel version check : Configure the Universal Profiling Agent to bypass the kernel version compatibility check.
- Environment variables for the Universal Profiling Agent : Configure the Universal Profiling Agent using the environment.
Command line arguments to the Universal Profiling Agent take precedence over environment variables, and environment variables take precedence over the configuration file.
The Universal Profiling Agent accepts the following CLI arguments:
CLI argument | Type | Example | Description |
---|---|---|---|
|
|
|
Run the Universal Profiling Agent in verbose mode. |
|
|
|
Disable the kernel version check. See Override kernel version check for more details. |
|
|
|
Set specific tags. See Tag data for querying for more details. |
|
|
|
Splits profiling data into logical groups that you control. You can assign any non-zero, unsigned integer ⇐ 4095. |
|
|
|
Set the secret token for communicating with the Universal Profiling Collector to |
|
|
|
Set the destination for reporting profiling information to |
|
|
|
Set the probabilistic interval to |
|
|
|
Set the probabilistic threshold to |
|
|
|
Set the path for the configuration file of the Universal Profiling Agent. See Configuration file of the Universal Profiling Agent for more details. |