Install CloudBurrow.

1Before you install

Docker, kind v0.33.0 and kubectl. Go only if you build from source. Give Docker at least 4 CPU and 6 GB; the full stack measures about 1.5 GiB and 2.1 CPU.

Windows is unsupported; WSL2 is untested. Container engines, with the evidence for each

2Install

brew install cloudburrow/tap/cloudburrowcloudburrow version

curl -fsSL https://raw.githubusercontent.com/cloudburrow/cloudburrow/main/scripts/install.sh | sh

git clone https://github.com/cloudburrow/cloudburrow.gitcd cloudburrowmake build

macOS binaries in v0.1.0 are not signed or notarized yet (#605). See Gatekeeper in docs/install.md.

3Check

cloudburrow doctor changes nothing, and fails only on what would stop up.

cloudburrow doctor  ok      docker                   /usr/local/bin/docker  ok      kind                     /opt/homebrew/bin/kind (v0.33.0)  ok      kubectl                  /usr/local/bin/kubectl  ok      docker daemon            29.8.0 (Docker Desktop)  ok      docker engine            Docker Desktop (in a VM)  ok      docker memory            39.1 GiB  ok      docker cpus              16  ok      disk space               24.2 GiB free on ~ (the host volume backing                                   the Docker Desktop VM disk, not the VM filesystem)  ok      port control             127.0.0.1:9000 is free  ...All checks passed.

ok

Passed.

warn

Will probably work: an untested kind release, or disk below the 20 GiB margin.

FAIL

up will not succeed. Exit code 1. Only FAIL blocks.

unknown

Could not be measured. Deliberately not ok, and does not block.

4Run

cloudburrow up

In another shell, point Google tooling at it:

eval "$(cloudburrow env)"

That exports endpoint overrides, the project, a local metadata server and a generated ADC file. The credentials authorise nothing; they exist so tooling that insists on them runs offline.

stop, reset and delete are distinct commands, and none implies another.

5Make your first call

import uuid
from google.cloud import storage

client = storage.Client()
bucket = client.create_bucket(f"example-{uuid.uuid4().hex[:12]}")
bucket.blob("hello.txt").upload_from_string("hello")
assert bucket.blob("hello.txt").download_as_text() == "hello"
bucket.delete(force=True)

import assert from 'node:assert/strict';
import { randomUUID } from 'node:crypto';
import { Storage } from '@google-cloud/storage';

const apiEndpoint = process.env.STORAGE_EMULATOR_HOST;
delete process.env.STORAGE_EMULATOR_HOST;
const storage = new Storage({ apiEndpoint, projectId: process.env.GOOGLE_CLOUD_PROJECT });

const [bucket] = await storage.createBucket(`example-${randomUUID()}`);
await bucket.file('hello.txt').save('hello', { resumable: false });
const [data] = await bucket.file('hello.txt').download();
assert.equal(data.toString(), 'hello');
await bucket.deleteFiles({ force: true });
await bucket.delete();

Go is tested through CloudBurrow's own compatibility suite. Java, .NET, Ruby and PHP are untested (#718).

6gcloud and Terraform

cloudburrow gcloud-setup configures the Google Cloud CLI. Verified command families: storage, kms, secrets, tasks, scheduler, logging and projects list. gcloud pubsub is partial; gcloud run is untested.

cloudburrow terraform sets endpoints only for Verified services and names the enabled services it left out. Verified with Terraform's hashicorp/google 8.4.0 and OpenTofu 1.12.6.

Every tool and the test behind it, in docs/status.md

7How long does up take?

One measured run on Ubuntu, 2026-09-24: the default profile started cold in 85 s and warm in 30 s. It is a single measurement, not a benchmark.

Source: Footprint, in docs/status.md

8Offline next release

cloudburrow prefetch downloads what up needs; up --offline then runs without a network.

Offline and air-gapped use, in docs/install.md