Configure the Container
Use the following steps to stand up the Private AI Services Container and connect it to the database.
- Navigate to the setup directory on the GPU host.
Note:
Ensure that you have copied the bash scripts included in the container image before completing these steps. The scripts are used to simplify container configuration. For instructions on how to copy the scripts, see Configure the Private AI Services Container.
The
setupdirectory used here is created as a result of unzipping the.zipfile containing the helper scripts. It is not the folder that the.zipwas copied into.cd setup -
Export helper variables used by the setup scripts.
export PRIVATE_DIR=$PWD/privateai export SECRETS_DIR=$PWD/secrets -
Create working directories for container assets and secrets.
mkdir -p "$PRIVATE_DIR" "$SECRETS_DIR" -
Generate or update the keystore that will hold TLS material.
./secretsSetup.sh -s "$SECRETS_DIR" -
Populate the container configuration and map UIDs and GIDs.
./configSetup.sh -d "$PRIVATE_DIR" -s "$SECRETS_DIR" -
Launch the container with the default image and HTTPS enabled.
./containerSetup.sh -d "$PRIVATE_DIR"For information about using the
--gputag to target a particular GPU, see Advanced Configuration Options. -
Gather the information required for configuring the database server.
echo "offload_url : https://$(hostname -f):8443/v1/index" echo "API key : $(cat $SECRETS_DIR/api-key)" echo "certificate : $SECRETS_DIR/cert.pem"Note:
The API key is equivalent to the database credential.
Parent topic: Configure the Vector Index Service