Infrastructure Management

How to provision and manage Cloudflare® D1 databases, KV namespaces, R2 buckets, Queues, and Vectorize indexes using hoox infra.

This page

Hoox is fully serverless, using Cloudflare®’s distributed services as its compute and storage engine. Managing these resources—such as provisioning D1 databases, registering KV config buckets, setting up S3-compatible R2 storage, and mounting asynchronous Queues—is done directly via the hoox infra command group (prefer this over manual dashboard / Wrangler calls).

This guide is your operations manual for provisioning, inspecting, and managing Hoox edge infrastructure.

⚡ 1. One-Click Quick Provisioning

When setting up your trading workspace for the first time, you do not need to manually configure resources in Cloudflare®’s dashboard. The Hoox CLI automatically parses your enabled workers' configurations and spins up the entire infrastructure stack in one command:

# Auto-provision all required KV, D1, R2, Vectorize, and Queue resources
hoox infra provision

This command performs a dependency audit and calls Cloudflare®'s APIs to provision:

  1. D1 SQLite Database (hoox-db).

  2. CONFIG_KV Namespace bucket.

  3. trade-execution

    messaging Queue.

  4. trade-reports

    and

    system-logs

    R2 object storage buckets.

  5. rag-index

    Vectorize vector search database.

🗄️ 2. D1 Database Administration

D1 databases store your critical transactional data (ledger, positions, balance history).

# A. List all active D1 databases in your Cloudflare® account
hoox infra d1 list

# B. Create a new custom database instance
hoox infra d1 create my-custom-db

# C. Delete a database instance (destructive)
hoox infra d1 delete my-custom-db

Note

Once a D1 database is created, the CLI will output its unique database_id (a 36-character UUID). You must ensure this ID is bound in your worker's wrangler.jsonc file so the worker V8 isolate can establish a connection at runtime.

🔑 3. KV Configuration Namespace Management

KV stores are used to manage fast-path configurations, global parameters, and the kill switch.

# A. List all KV namespaces in your account
hoox infra kv list

# B. Create a new KV namespace (e.g. for staging or production)
hoox infra kv create CONFIG_KV

# C. Apply the default 16-key configuration manifest to your active namespace
hoox config kv apply-manifest

📨 4. Asynchronous Queues Setup

Queues guarantee order delivery during periods of extreme exchange rate limits or network congestion.

# A. List all active Cloudflare® Queues
hoox infra queues list

# B. Create the trade execution queue
hoox infra queues create trade-execution

Queue Parameters & Throttling

During provisioning, Hoox configures the queue with:

  • Retry Backoff: 30 seconds initial delay, scaling exponentially up to 15 minutes.
  • Message Retention: 4 days (ensuring messages are preserved even during prolonged exchange maintenance events).

📦 5. R2 Object Storage Administration

R2 is an S3-compatible, zero-egress fee object storage service used to offload heavy JSON payloads and PDF portfolio reports.

# A. List all R2 buckets
hoox infra r2 list

# B. Create the trade-reports bucket
hoox infra r2 create trade-reports

🧠 6. Vectorize RAG Index Management

Vectorize indexes house high-dimensional vector embeddings that power semantic search and Telegram AI bot memory.

# A. List all Vectorize indexes
hoox infra vectorize list

# B. Create a vector index with specific dimensions (e.g. 1536 for OpenAI embeddings)
hoox infra vectorize create rag-index --dimensions 1536 --metric cosine

Warning

Destroying resources via hoox infra <service> delete is highly destructive and irreversible. Deleting a D1 database instantly wipes your trading records; deleting a KV bucket drops all configurations and triggers immediate gateway errors. Always backup ledgers before running deletions!

🔗 Next Steps