Blockchain Node Infrastructure Setup Guide

You are a blockchain DevOps engineer who manages node infrastructure for DeFi protocols and data providers. You have experience running full nodes, archive nodes, and specialized infrastructure across Ethereum, Solana, and L2 networks.

CONTEXT: My DeFi protocol needs reliable blockchain infrastructure. We currently use Alchemy and Infura for RPC access, but as we scale, we need to consider running our own nodes for cost savings, lower latency, and independence from third-party providers. We also need archive node access for historical data queries. Our current RPC usage is approximately 50 million requests per month.

TASK: Create a blockchain node infrastructure guide:
1. Infrastructure options comparison: evaluate the approaches — third-party RPC providers (Alchemy, Infura, QuickNode, Ankr), self-hosted nodes (dedicated hardware or cloud VMs), decentralized RPC networks (Lava, dRPC, Pocket Network), and hybrid approaches. Compare by: cost at different usage levels, latency, reliability, customization, and data access (full vs. archive).
2. Self-hosted node setup: guide for running Ethereum nodes — hardware requirements for full node vs. archive node (CPU, RAM, storage — archive needs 12+ TB SSD), execution client options (Geth, Nethermind, Besu, Erigon — compare by performance and resource usage), consensus client pairing (Lighthouse, Prysm, Teku), and sync time estimates. Include the complete setup process.
3. L2 node infrastructure: how to run nodes for major L2s — Arbitrum (Nitro node), Optimism (op-node + op-geth), and Base (OP Stack node). Specify additional hardware requirements, sync process, and the relationship between L2 nodes and L1 data.
4. Cost analysis: compare monthly costs — self-hosted (cloud: AWS/GCP/Hetzner pricing for required instances, bandwidth costs) vs. managed providers (Alchemy Growth plan, QuickNode Build plan) vs. decentralized networks. Calculate break-even point where self-hosting becomes cheaper than managed providers. Include hidden costs (DevOps time, monitoring tools, redundancy).
5. High-availability architecture: design a production node setup — load balancing across multiple nodes, failover configuration, health monitoring and alerting (Prometheus, Grafana, PagerDuty), geographic distribution for latency optimization, and disaster recovery procedures.
6. Data indexing infrastructure: beyond RPC access — when to deploy a subgraph (The Graph) vs. custom indexer, Ponder and Envio as alternatives to The Graph, and database considerations for storing indexed data (PostgreSQL, TimescaleDB for time-series). Design the indexing stack for a lending protocol (events to index, query patterns).