Networks Fukuii runs
The production networks ship configured, each with its testnet: Ethereum Classic, the only Proof-of-Work EVM network in production, and Ethereum, the largest Proof-of-Stake network.
They are grouped into families rather than listed flat, because the difference between them is structural rather than a matter of settings. A Proof-of-Work family dispatches its forks by block number and has no deposit contract; a Proof-of-Stake family dispatches by timestamp and does. Grouping them means the shared code never asks which network it is running, and adding a family does not mean adding a branch to everything that already works.
Ethereum Classic
Ethereum
PoA & consortium networks
Clique/Qbft/Bor-style consensus, custom genesis and fork schedule, no source changes
Sidechains
Polygon/Bor-style, with injected bridge/oracle logic
Development networks
Generate a genesis and stand up a private or test network from the same binary
L2 rollups
OP-stack/Taiko-style, including rollup-specific transaction types and derivation
ZK-EVM & alt-consensus
ZK-EVM and alt-BFT alongside PoW and PoS, including custom VMs and zk-gas schedules
Each family is selected per deployment. Archival storage and snapshot distribution are covered on the operations page.
What makes one network different
A network is a set of values, not a build. These are what a configuration supplies, and pointing Fukuii at a different chain means supplying a different set rather than compiling a different client.
| What | Why it is separate |
|---|---|
| Chain ID | Signs transactions, so it decides what a signature is valid on |
| Network ID | Gates peering, so it decides who a node will talk to |
| Genesis hash | Identifies the data directory, so it decides which chain a database holds |
| Fork schedule | By block number or by timestamp, depending on the family |
| Gas schedule | Fee mechanics and any divergence from mainnet pricing |
| Consensus driver | What decides which block is canonical |
The first three are independent on purpose. Chain ID signs transactions, network ID gates peering, and the genesis hash identifies a data directory, so a custom network can share one with another chain and differ on the rest. Treating them as a single identity is a common way to end up with a node that peers happily and then rejects every transaction it receives.
What running networks concurrently means
Fukuii supervises each network as a separate unit. The networks share a process but not their data, so one network's sync cannot affect another's.
state
Each network keeps its own storage. Nothing is shared, so one chain's sync cannot corrupt or bloat another's database.
metrics
Each network gets its own metrics registry, so a dashboard reads one chain rather than an aggregate of several.
config
Fork schedule, peers, RPC surface and consensus module are all set per network, in one process with one supply chain to audit.
The network list above is what ships configured. The consensus module it sits on is a seam, so a deployment can point it at a consortium chain, a sidechain, or an L2 without a different build.
Permissioned and private networks →