Files
beanflows/infra/__main__.py
Deeman 719aa8edd9 Remove R2 bucket management from Pulumi, use cpx11 for supervisor
- R2 buckets (beanflows-artifacts, beanflows-data-prod) managed manually in Cloudflare UI
- R2 API tokens don't work with Cloudflare Pulumi provider
- Use cpx11 (€4.49/mo) instead of non-existent ccx11
- Import existing SSH key (deeman@DeemanPC)
- Successfully deployed supervisor at 49.13.231.178
2025-10-12 23:18:52 +02:00

117 lines
3.8 KiB
Python

"""
BeanFlows.coffee Infrastructure
Hetzner compute stack for ephemeral worker orchestration
Note: R2 buckets are managed manually in Cloudflare dashboard
- beanflows-artifacts: Stores CLI and pipeline artifacts
- beanflows-data-prod: Iceberg data lakehouse
"""
import pulumi
import pulumi_hcloud as hcloud
# Load configuration
config = pulumi.Config()
hetzner_location = config.get("hetzner_location") or "nbg1" # Nuremberg datacenter
# ============================================================
# R2 Bucket Names (managed manually in Cloudflare R2 UI)
# ============================================================
# R2 buckets cannot be managed via Pulumi as they require R2-specific tokens
# that don't work with the Cloudflare Pulumi provider.
# These are defined here for documentation purposes only.
ARTIFACTS_BUCKET = "beanflows-artifacts" # CLI + extract/transform packages
LAKEHOUSE_BUCKET = "beanflows-data-prod" # Iceberg tables (EEUR region)
# ============================================================
# Hetzner Cloud Infrastructure
# ============================================================
# SSH key for server access (imported from existing key)
ssh_key = hcloud.SshKey(
"materia-ssh-key",
name="deeman@DeemanPC",
public_key=config.require_secret("ssh_public_key"),
opts=pulumi.ResourceOptions(protect=True),
)
# Small CPX instance for supervisor (runs materia CLI to orchestrate pipelines)
# This is an always-on instance that creates/destroys ephemeral workers on-demand
supervisor_server = hcloud.Server(
"materia-supervisor",
name="materia-supervisor",
server_type="cpx11", # 2 vCPU (shared), 2GB RAM, ~€4.49/mo (cheapest option)
image="ubuntu-24.04",
location=hetzner_location,
ssh_keys=[ssh_key.id],
labels={
"role": "supervisor",
"project": "materia",
},
user_data="""#!/bin/bash
set -e
# Basic server setup
apt-get update
apt-get install -y python3.13 python3-pip curl unzip
# Install Pulumi ESC CLI
curl -fsSL https://get.pulumi.com/esc/install.sh | sh
export PATH="$HOME/.pulumi/bin:$PATH"
echo 'export PATH="$HOME/.pulumi/bin:$PATH"' >> /root/.bashrc
# Create deployment directory
mkdir -p /opt/materia
# Configure environment
echo 'Setup complete. Materia CLI will be deployed via CI/CD.' > /opt/materia/README.txt
""",
)
# Note: Workers are created on-demand by the materia CLI
# No always-on worker instances in this architecture
# Firewall for servers (restrict to SSH + outbound only)
firewall = hcloud.Firewall(
"materia-firewall",
name="materia-firewall",
rules=[
# Allow SSH from anywhere (consider restricting to your IP)
hcloud.FirewallRuleArgs(
direction="in",
protocol="tcp",
port="22",
source_ips=["0.0.0.0/0", "::/0"],
),
# Allow all outbound traffic
hcloud.FirewallRuleArgs(
direction="out",
protocol="tcp",
port="any",
destination_ips=["0.0.0.0/0", "::/0"],
),
hcloud.FirewallRuleArgs(
direction="out",
protocol="udp",
port="any",
destination_ips=["0.0.0.0/0", "::/0"],
),
],
)
# Apply firewall to supervisor
supervisor_firewall = hcloud.FirewallAttachment(
"supervisor-firewall",
firewall_id=firewall.id,
server_ids=[supervisor_server.id],
)
# ============================================================
# Outputs
# ============================================================
pulumi.export("supervisor_ip", supervisor_server.ipv4_address)
pulumi.export("artifacts_bucket_name", ARTIFACTS_BUCKET)
pulumi.export("lakehouse_bucket_name", LAKEHOUSE_BUCKET)