Gnoland Mainnet Snapshot

POSTHUMAN publishes a verified LZ4 snapshot every six hours from a dedicated non-signing gnoland-1 archive RPC node.

• Latest archive: https://snapshots-gnoland.posthuman.digital/gnoland_gnoland-1-latest.tar.lz4
• SHA-256: https://snapshots-gnoland.posthuman.digital/SHA256SUMS
• Metadata: https://snapshots-gnoland.posthuman.digital/snapshot.json

The archive contains only db/ and wal/. It never contains secrets/, configuration, genesis, or validator signer state, so it can be restored into a node of any role without touching identity.

snapshot.json states the exact height and the app hash of the snapshot, which is what makes an archive verifiable before you trust it.

Restore

Set the data directory used by your service:

GNO_HOME="$HOME/gnoland-mainnet"
DATA_DIR="$GNO_HOME/data"
SNAPSHOT_URL=https://snapshots-gnoland.posthuman.digital
SNAPSHOT_FILE=gnoland_gnoland-1-latest.tar.lz4
TMP_DIR=$(mktemp -d)
BACKUP_DIR="$HOME/gnoland-backup-$(date -u +%Y%m%dT%H%M%SZ)"

Download and verify before touching the running node — the download is the long part, and a node that is still running loses no blocks meanwhile:

curl -fL --retry 3 -o "$TMP_DIR/$SNAPSHOT_FILE" "$SNAPSHOT_URL/$SNAPSHOT_FILE"
curl -fL --retry 3 -o "$TMP_DIR/SHA256SUMS" "$SNAPSHOT_URL/SHA256SUMS"

cd "$TMP_DIR"
sha256sum -c SHA256SUMS
lz4 -t "$SNAPSHOT_FILE"
if lz4 -dc "$SNAPSHOT_FILE" \
  | tar -tf - \
  | grep -Evq '^(db|wal)(/|$)'; then
  echo "Unexpected snapshot path"
  exit 1
fi

Only now stop the service and back up identity, configuration, and genesis:

sudo systemctl stop gnoland.service
install -d -m 0700 "$BACKUP_DIR"
cp -a "$DATA_DIR/secrets" "$DATA_DIR/config" "$DATA_DIR/genesis.json" \
  "$BACKUP_DIR/"

Preserve the old database and restore the new one:

mv "$DATA_DIR/db" "$BACKUP_DIR/db.old"
mv "$DATA_DIR/wal" "$BACKUP_DIR/wal.old"
lz4 -dc "$TMP_DIR/$SNAPSHOT_FILE" | tar -xf - -C "$DATA_DIR"
sudo systemctl start gnoland.service

Verify chain, progress, and service health:

systemctl is-active gnoland.service
curl -fsS http://127.0.0.1:26657/status | jq '.result | {
  network: .node_info.network,
  height: .sync_info.latest_block_height,
  catching_up: .sync_info.catching_up,
  voting_power: .validator_info.voting_power
}'
sudo journalctl -u gnoland.service --since '10 minutes ago' --no-pager

The reported network must be gnoland-1. Keep the old db/ and wal/ until the restored node is synced and stable. For a validator, independently verify recent signatures before deleting the rollback copy.

AppHash mismatch recovery

If the node crashes with an AppHash mismatch, preserve logs and the current db/ + wal/ for forensics. Restore only those directories from a known healthy backup or this verified snapshot. Never overwrite validator keys, node identity, configuration, genesis, or signer state.