Daily snapshots are published for every network. Restoring from the latest pruned snapshot is the fastest way to sync a node — minutes instead of replaying the full chain.

The panel above reads the live catalog (pruned/snapshot.json) straight from the bucket. If it fails to load, the URLs below follow the same layout.

Download and restore (mainnet)

CHAIN_ID="morocco-1"
BASE="https://s3.terp.network/snapshots/mainnet/${CHAIN_ID}"
 
# Resolve the latest published snapshot
LATEST_URL=$(curl -s ${BASE}/pruned/snapshot.json | jq -r .latest)
echo "Latest: ${LATEST_URL}"
 
# Stop the node, clear state, fetch and extract
sudo systemctl stop terpd
rm -rf $HOME/.terpd/data $HOME/.terpd/wasm
curl -sL "${LATEST_URL}" | lz4 -dc | tar -x -C $HOME/.terpd
 
# Restart and let it catch up to head
sudo systemctl start terpd
terpd status | jq .sync_info.catching_up

Use https://minio.terp.network/snapshots/... as an alternate host for the same paths.

Testnet (120u-1) does not publish a daily pack — sync testnet from genesis plus state sync instead.

Each tarball is named {chain_id}_{height}_{UTC}Z.tar.lz4 and bundles data/ plus wasm/. The height is right in the filename — confirm it is recent before restoring:

curl -s ${BASE}/pruned/snapshot.json | jq -r .latest
# .../morocco-1_22945782_2026-09-03T13-03-09Z.tar.lz4  →  height 22945782

Snapshot tiers

terpd snapshot produces local snapshots in tiers; the published daily pack is the pruned tier:

Archive packs (full history) are published under archive/ on the same layout when available. See State sync for the no-download alternative and Archive for running a history node.

Roll your own

To snapshot your own node instead of downloading, stop it first so data/ is consistent, then archive:

sudo systemctl stop terpd
tar -C $HOME/.terpd -cf - data wasm | lz4 -c > $HOME/morocco-1_$(date '+%Y-%m-%d').tar.lz4
sudo systemctl start terpd

Never restore a snapshot onto a running node, and never copy a live data/ directory without stopping the process — you will get a corrupt state.