Running AI Gateway as a standalone binary¶
Before you begin¶
To run the F5 AI Gateway binary, you will need an environment running Linux.
To run F5 AI Gateway, you need one or more processors and one or more pre-configured services. Review the related sections in the documentation to learn more.
If you are running any processors or services locally, you need to install and configure Docker (or your container platform of choice), then install and configure the processors and services that you plan to use.
Using the standalone binary¶
The standalone binary is a single executable file that can be run on any machine with the required dependencies. During this initial phase, the F5 customer success team will provide you with the binary file.
To start the AI Gateway core, run the standalone binary file aigw
with one of the following commands:
To start without additional arguments::
aigw start
To start with a configuration file::
aigw start </path-to/config.yaml>
Note
Starting the AI Gateway without arguments will use the default configuration. The admin server will use the default port, 8080.
Flags¶
The following optional flags are available when running the aigw start
command:
Flag |
Description |
Default |
---|---|---|
|
Allows |
|
|
The directory where audit data will be stored. Must be used in conjunction with the |
System’s default cache folder + |
|
Configures the batch size of the exporter. The Exporter will not export closed transactions until their amount is greater or equal than the batch size. |
|
|
Enable exporting transactions. |
|
|
Removes transactions after exporting. |
|
|
Sets the name of an S3 bucket where exporter will upload transactions. |
|
|
Sets the timeout for uploading a single transaction. |
|
|
Configures the type of the exporter, must be one of: [stdout s3]. |
|
|
Configures the number of workers (goroutines) for the exporter. |
|
Configuration file¶
The configuration file is a YAML file that defines the AI Gateway configuration.
To learn more about the configuration file, see the
[configuration file])configuration-file) section of the documentation. You can
find an example configuration file in the root of the AI Gateway repository,
(aigw.yaml
).