Network node
A named client and network, with chain-state storage, synchronisation, update, snapshot, peer and recovery procedures.
Beyond mining
Blockchain infrastructure starts with the network role.
A node, RPC endpoint, indexer, explorer feed or development environment has different storage, availability, security and maintenance needs. Specify the service before choosing the server.
Possible service patterns
One label, several very different systems.
These are scoping patterns rather than advertised managed services. Availability and support commitments require a separate contract.
RPC and indexing
Controlled application access, rate limits, logs, database growth, query requirements and a measured availability target.
Development environment
Isolated test networks, deployment tooling, key-handling boundaries, CI inputs and clear separation from production assets.
Discovery record
Inputs needed before a specification.
- Named blockchain, client and network role
- Expected chain state, history and growth
- Read, write and indexing workload
- Users, applications and authentication route
- Availability, backup and recovery objectives
- Monitoring, patching and incident owner
No custody implied
Infrastructure support must not silently become custody of private keys, signing authority or customer funds. Key generation, storage, signing and recovery need an explicit owner and approved procedure.
Hardware fit
Not every blockchain workload needs a GPU.
Storage latency, database performance, network throughput and memory may matter more than accelerator count. GPU capacity belongs in the design only when the named workload can use it.