Deploy to Vespa Cloud

This notebook deploys a Vespa.ai sample application to Vespa Cloud using pyvespa. The Vespa CLI is used for managing security credentials.

See pyvespa, and jupyter notebooks requirements to run this notebook.

Refer to the sample apps site for more applications.

Install pyvespa and the Vespa CLI

!pip install pyvespa

If homebrew is installed, install the Vespa CLI:

!brew install vespa-cli

Alternatively, when running in Colab, install the Vespa CLI:

import os
import requests
res = requests.get(url="https://api.github.com/repos/vespa-engine/vespa/releases/latest").json()
os.environ["VERSION"] = res["tag_name"].replace("v", "")
!curl -fsSL https://github.com/vespa-engine/vespa/releases/download/v${VERSION}/vespa-cli_${VERSION}_linux_amd64.tar.gz | tar -zxf -
!ln -sf /content/vespa-cli_${VERSION}_linux_amd64/bin/vespa /usr/local/bin/vespa

Create a Vespa Cloud tenant

Sign up and create a tenant: Run steps 1 and 2 in getting started, then replace the tenant name below:

import os
os.environ["TENANT_NAME"] = "mytenant"

Set up security credentials

Configure the tenant.application.instance:

!vespa config set target cloud
!vespa config set application ${TENANT_NAME}.myapp.default

If the tenant.application.instance is already created with security credentials, skip next two steps, and go directly to setting api_key_path. If not, create data-plane credentials:

!vespa auth cert -N

vespa auth cert -N stores the credentials in ~/.vespa, like:

/Users/me/.vespa/mytenant.myapp.default/data-plane-public-cert.pem
/Users/me/.vespa/mytenant.myapp.default/data-plane-private-key.pem

Create an api-key, used to deploy the application:

!vespa auth api-key

vespa auth api-key stores the api key in ~/.vespa, like:

/Users/me/.vespa/mytenant.api-key.pem

Add it in the console per instructions from the command. Enter full path to the api key:

api_key_path = "/Users/me/.vespa/mytenant.api-key.pem"

Create an application package

An application package is the application’s configuration, like schema. Here, using a basic question answering app from the app gallery.

from vespa.gallery import QuestionAnswering

app = QuestionAnswering()

Create a VespaCloud instance

Refer to the VespaCloud reference.

from vespa.deployment import VespaCloud

vespa_cloud = VespaCloud(
    tenant=os.getenv("TENANT_NAME"),
    application="myapp",
    key_location=api_key_path,
    application_package=app)

Deploy to Vespa Cloud

Refer to tenants, applications and instances for details - here creating an instance named default:

app = vespa_cloud.deploy(instance="default")

That is it, you can now interact with your deployed application through the app instance.