results

README

Tekton Results API

This package contains experimental code to support a richly queryable API for Tekton execution history and results.

The full proposal is here: https://docs.google.com/document/d/1-XBYQ4kBlCHIHSVoYAAf_iC01_by_KoK2aRVO0t8ZQ0/edit

The main components of this design are a queryable indexed API server backed by persistent storage, and an in-cluster watcher to report updates to the API server.

The API server interface is defined in ./proto/api.proto, and a reference implementation backed by Sqlite will live in ./cmd/api. A reference implementation of the in-cluster watcher will live in ./cmd/watcher.

Development

Configure your database.

The reference implementation of the API Server requires a SQL database for result storage. The database schema can be found under schema/results.sql.

Initial one-time setup is required to configure the password and initial config:

kubectl create secret generic tekton-results-mysql --namespace="tekton-pipelines" --from-literal=user=root --from-literal=password=$(openssl rand -base64 20)
kubectl create configmap mysql-initdb-config --from-file="schema/results.sql" --namespace="tekton-pipelines"

Deploying

To build and deploy both components, use ko. Make sure you have a valid kubeconfig, and have set the KO_DOCKER_REPO env var.

ko apply -f config/

To only build and deploy one component:

ko apply -f config/watcher.yaml

Regenerating protobuf-generated code

1. Install protobuf compiler

e.g., for macOS:

brew install protobuf

1. Install the protoc Go plugin

$ go get -u github.com/golang/protobuf/protoc-gen-go

1. Rebuild the generated Go code

$ go generate ./proto/