DiggyDB - Amazon Route53 as a blazingly fast and reliable database 🚀

🚧 This is currently a work in progress, and subject to change. Yes, it's a serious thing too...

This one is 100% inspired by Corey Quinn, who has repeatedly insisted that Route53 is essentially a database with 100% SLA.

DiggyDB goes one step further than simple a key/value TXT record by allowing you to use (or indeed abuse) DNS TXT records by storing JSON data, almost as though it was a MongoDB or AWS DynamoDB!

JSON data is transformed and stored in a TXT record with each key/value pair on a seperate line:

"name=gatsby"
"version=3.1.2"
"downloads=409,484"

Querying for a known id (via DNS) is blazingly fast and can take ~30ms to return the JSON data. 🚀

Features

Query table by id (via DNS)
Query all rows in a table (via AWS SDK)
Put a new object in a table (via AWS SDK)

More to come soon, but feel free to contribute if you have ideas!

Limitations

Due to the nature of TXT records, object values should be limited to alphanumeric string values.

Some DNS servers may ignore the TTL value of DNS records, so updates may be delayed.

Install

npm install -S diggydb-nodejs

or

yarn add diggydb-nodejs

How to use

First you need to import diggydb-nodejs and configure it:

import { DiggyDB, DiggyQuery } from "diggydb-nodejs"

const db = new DiggyDB({
  hostname: "example.com",
  accessKeyId: process.env.AWS_ACCESS_KEY_ID,
  secretAccessKey: process.env.AWS_SECRET_ACCESS_KEY,
})

Query a row by id

If you know the id of the row you would like to retrieve from a table, you can use the following:

interface Framework {
  id: string
  name: string
  version: string
  downloads: string
}

const id = "3a7b68cf-686a-4135-b48b-6471deb3643a"
const table = "frameworks"

const result = await db.query<Framework>(id, table)

// returns
const result = {
  id: "3a7b68cf-686a-4135-b48b-6471deb3643a",
  name: "gatsby",
  version: "3.1.2",
  downloads: "409,484",
}

Query rows in a table

You can retrieve all rows from a table using the following:

interface Framework {
  id: string
  name: string
  version: string
  downloads: string
}

const table = "frameworks"

const results = await db.queryTable<Framework[]>(table)

// returns
const results = [
  {
    id: "3a7b68cf-686a-4135-b48b-6471deb3643a",
    name: "gatsby",
    version: "3.1.2",
    downloads: "409,484",
  },
  {
    id: "aae3d92e-7230-43e4-b63c-1efe00081ae6",
    name: "next.js",
    version: "10.0.9",
    downloads: "1,144,043",
  },
]

Adding a row to a table

You can add a row to a table using the following:

import { v4 as uuidv4 } from "uuid"

interface Framework {
  id: string
  name: string
  version: string
  downloads: string
}

const table = "frameworks"
const id = uuidv4() // create a unique id

const newRecord: Omit<Framework, "id"> = {
  name: "vuejs",
  version: "2.6.12",
  downloads: "1,935,726",
}

await db.put(table, uuidv4(), newRecord)

Note: If the table doesn't exist, it will of course create it for you

Adding records manually (or not using AWS)

It's not just Amazon Route53 you can use as a database, but you would need to update your DNS records manually:

For the record type, select TXT.
In the Name/Host/Alias field, enter id.table.diggydb. Your host might require you to enter the fully qualified domain, so this may look like id.table.diggydb.example.com. Your other DNS records might indicate what you should enter.
In the Time to Live (TTL) field, enter 60 or a value of your choice. This is time in seconds that DNS servers should cache the record for.
In the Value/Answer/Destination field, enter your JSON data key/value pairs (excl the id field) on each line:

"key=value"
"key=value"
"key=value"

Save the record.

njbmartin/diggydb-nodejs