Preset configs

Six ready-to-use configs ship in config-examples/. Each is a small JSON file you can copy and adjust if needed.

Pick one based on what you'll query, not on what looks comprehensive — disabling projections you don't need is the biggest single lever on sync time and disk usage.

Preset	Projections enabled	Ledger	Approximate size (mainnet)
minimal.json	(core only)	off	~30 GB
utxo-only.json	utxo	off	~160 GB
spo.json	utxo, metadata, stake_delegation, pool	off	~185 GB
dapp.json	utxo, multi_asset, metadata, scripts_datums	off	~265 GB
everything-no-ledger.json	all non-ledger projections	off	~475 GB
everything.json	every projection, ledger on	on	~480 GB

The everything figure is measured at mainnet tip; the smaller-preset figures are approximations derived from its per-table breakdown.

Every preset also enables epoch_sync_stats so the per-epoch sync progress table gets populated. It's cheap and useful for diagnostics.

The cbor tax

tx_cbor (raw transaction CBOR bytes) accounts for ~218 GB on its own at mainnet tip — nearly half of an everything-profile database. Both everything-no-ledger.json and everything.json enable cbor by default. If your consumers don't need to re-serialise or replay transactions, copying one of those presets and setting "cbor": false cuts the database from ~475 GB to ~260 GB.

minimal

Block / tx index only. The core extractor writes block, tx, and slot_leader rows; no UTxO, no certificates, no metadata. Suitable for:

• Hash-to-height resolution.
• Per-block summary queries (block count per epoch, slot leader distribution).
• Verifying that your dbsync stack is wired up end-to-end before committing to a larger profile.

It's the fastest possible sync — useful as a sanity check on new hardware. Not useful for any wallet, balance, or asset query.

utxo-only

The utxo projection on top of core. Adds tx_out, tx_in, collateral_tx_in, collateral_tx_out, reference_tx_in, and address.

Suitable for:

• Wallet backends — resolve an address to its current UTxO set.
• Balance queries — sum unconsumed outputs for an address.
• Spend-tracing — tx_in.tx_out_id resolves to the producing output.

consumed_by_tx_id is enabled by default (populates a back-pointer from each tx_out to the tx that spent it). tx_in is enabled by default. The archive strategy stores every output ever produced; the prune and from_ledger strategies are reserved but not yet implemented.

spo

Stake-pool operator view. Adds metadata, stake_delegation, and pool on top of utxo.

Tables it gives you:

• stake_address, stake_registration, stake_deregistration, delegation, withdrawal.
• pool_hash, pool_update, pool_metadata_ref, pool_owner, pool_relay, pool_retire.
• tx_metadata for general tx metadata queries.

Suitable for:

• Pool dashboards — list of registered pools, current delegation, retirement schedule.
• Tracking a specific stake address's delegation history.
• Reward-address resolution.

Ledger is off, so reward amounts and deposit values aren't in the database. If you need those, use the everything preset instead.

dapp

DApp / explorer view. Adds multi_asset, metadata, and scripts_datums on top of utxo.

Tables it gives you:

• multi_asset, ma_tx_mint, ma_tx_out for native tokens.
• tx_metadata for arbitrary on-chain metadata.
• datum, script, redeemer, redeemer_data, extra_key_witness for Plutus and native-script witness data.

Suitable for:

• NFT / token explorers.
• DApps that need on-chain metadata payloads.
• Plutus contract analytics.

everything-no-ledger

The broad projection set that doesn't need ledger state. Adds stake_delegation, pool, governance, and cbor on top of dapp. It deliberately leaves out the ledger-derived projections (stake_delegation_ledger, pool_stats, epoch_boundary) and the off-chain metadata fetchers (off_chain_pools, off_chain_votes).

Suitable for:

• Block-explorer backends that need the full on-chain schema without reward / stake-distribution data.
• Downstream consumers that re-serialise transactions (the cbor extractor stores raw tx bytes).

governance records proposals, votes, and DRep / committee certificates straight from transaction bodies, so it works here without ledger. Only its drep_distr (per-epoch DRep voting power) stays empty, since that's derived from ledger state.

everything

ledger.enabled = true plus every projection: everything in everything-no-ledger, and additionally the ledger-derived stake_delegation_ledger, pool_stats, and epoch_boundary, the off_chain_pools / off_chain_votes metadata fetchers, and the reserved current_state stub.

With ledger on, the consumer applies blocks to an in-RAM ledger state and the worker produces:

• Per-block deposit maps (tx-level deposit amounts).
• Per-epoch ada_pots, epoch_param, epoch_state, reward, pot_reward, epoch_stake, and pool_stat.

This is the largest, slowest profile, and the one that gives you the full upstream-cardano-db-sync-equivalent schema.

Switching profiles

You can't switch profiles in place — see Profile immutability. A switch means a fresh sync against a fresh database. Plan the profile choice before you commit to a multi-day sync.