Getting Started Guide

Setup

Create a basic go project

Make sure glide and $GOPATH/bin are in your $PATH:

Starting in the base directory of your sources:

export PATH="$PATH:$GOPATH/bin"

Install the DOSA library

glide get github.com/uber-go/dosa

Install the DOSA CLI

This is a hack for the time being, ideally the command above would install dosa into $GOPATH/bin

go get -u github.com/uber-go/dosa/cmd/dosa

The dosa binary should now be installed in $GOPATH/bin.

ls -l $GOPATH/bin/dosa

Start Cerberus to proxy to dosa-gateway

cerberus -t dosa-gateway --dc sjc1

Create your DOSA development scope

dosa scope create your_name_here

Build your data model

Modeling your data is beyond the scope of this quick start guide. Please reach out to us for help!

Create at least one entity

In your code, create an object that embeds (github.com/uber-go/dosa).Entity, we'll call it entities.go:

package main

import (
    "time"

    "github.com/uber-go/dosa"
)

type MyFirstDosaObject struct {
    dosa.Entity `dosa:"primaryKey=((K1, K2), K3)"`
    K1 dosa.UUID
    K2 string
    K3 int64
    C1 time.Time
}

This example described an object with a partitioning key of (K1, K2) and a clustering key of K3. There is a single non-key field, C1.

Dump your schema definitions

Do this to check for errors in your entities, use the CLI's schema dump command:

dosa schema dump

This command takes an optional list of directories to look for entities. If none are provided, it will default to the current directory. Since the example assumes entities are defined in ./entities.go, no arguments are needed.

Install your schema with the schema upsert command:

dosa schema upsert

This will translate your DOSA entities into the associated database schema. If the underlying datastore is Cassandra, this will translate to CQL. For a preview of the schema that will be installed, use the schema dump command can again be used. In this case, to preview the CQL pass the format option:

dosa schema dump --format cql

TODO: Initialize client

TODO: Create/query data