Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.novacula.io/llms.txt

Use this file to discover all available pages before exploring further.

Sui nodes run as a single sui-node process per node. The chain adapter renders one fullnode.yaml (nested-dict format), downloads the genesis blob from the official Sui repository on first start, and points the node at network-specific archive and P2P seed lists.

​
Networks and resources

NetworkCPURAMDisk
mainnet16 cores128 GiB4000 GiB
testnet8 cores64 GiB3000 GiB
Sui defaults are heavy because the network’s state grows fast and pruning is configurable rather than aggressive. Lower at your own risk; the wizard surfaces a warning under defaults.

​
Ports

NamePortNotes
JSON-RPC9000Public read API
Network (TCP)8080Inter-validator RPC
P2P (UDP, QUIC)8084Discovery / gossip
Metrics9184Prometheus exporter

​
Node types

Full only. Sui doesn’t expose a separate archive mode; for archival use cases adjust pruning via overrides (see below).

​
Init steps

On first reconcile per node:
  1. Genesis blob — downloaded from github.com/MystenLabs/sui-genesis/raw/main/<network>/genesis.blob and written to the data dir.
  2. P2P seeds — written into fullnode.yaml from the adapter’s network-specific seed map. These are pinned validator-network endpoints (e.g. mel-00.mainnet.sui.io, ewr-00.mainnet.sui.io).
  3. Archive ingestion URL — pointed at https://checkpoints.<network>.sui.io for state-sync from the Mysten-operated archive.

​
Config overrides

Overrides go into fullnode.yaml (nested dict). The schema declares:
  • Managed paths — anything controlling networking (network-address, metrics-address, json-rpc-address), data layout (db-path), genesis path, and seeds. Rejected at validate time.
  • Known keys — common tuning knobs the UI renders with type hints + descriptions:
KeyTypeDescription
enable-event-processingbooleanEnable event indexing/processing
authority-store-pruning-config.num-epochs-to-retainnumberEpochs of history to retain in the store
authority-store-pruning-config.num-epochs-to-retain-for-checkpointsnumberEpochs of checkpoints to retain
authority-store-pruning-config.periodic-compaction-threshold-daysnumberDays between RocksDB compactions
state-archive-read-configstringOverride archive ingestion-url config (advanced)
Free-form keys not on the managed list are accepted as-is. Sui doesn’t use CLI arg overrides; everything flows through fullnode.yaml.

​
Metrics

Scraped from the sui-node Prometheus exporter on :9184. Surfaced:
  • Current checkpoint, current epoch.
  • Peer count.
  • Process CPU + memory, disk usage.
See Node monitoring.

​
Common patterns

  • Public RPC node — accept defaults; expose :9000 through your own reverse proxy.
  • Long-history node — bump num-epochs-to-retain and num-epochs-to-retain-for-checkpoints substantially; expect proportional disk growth.
  • Lean node (events off) — set enable-event-processing: false; the node won’t index events but uses materially less disk.

​
Performance note

Sui benefits more from fast NVMe than from raw CPU count — the bottleneck is RocksDB. If your storage class doesn’t deliver consistent low-latency writes (cloud burst-credit volumes, network filesystems), expect periodic stalls regardless of CPU/RAM headroom.