Persistence Node Snapshot

Our Persistence Snapshot Server Setup

We take a node snapshot every day. We then delete all the previous snapshots to free up the space on the file server.

The snapshot is designed for node opeartors to run an efficient validator service on Persistence chain. To make the snapshot as small as possible while still viable as a validator, we use the following setting to save on the disk space. We suggest you make the same adjustment on your node too. Please notice that your node will have very limited functionality beyond signing blocks with the efficient disk space utilization. For example, your node will not be able to serve as a RPC endpoint (which is not suggested to run on a validator node anyway).

Since we periodically state-sync our snapshot nodes, you might notice that sometimes the size of our snapshot is surprisingly small.

app.toml

# Prune Type
pruning = "custom"

# Prune Strategy
pruning-keep-recent = "100"
pruning-keep-every = "0"
pruning-interval = "10"

config.toml

indexer = "null"

How To Process Persistence Snapshot

Install lz4 if needed

sudo apt update
sudo apt install snapd -y
sudo snap install lz4

Download the snapshot

wget -O persistence_11674615.tar.lz4 https://snapshots.polkachu.com/snapshots/persistence/persistence_11674615.tar.lz4 --inet4-only

Stop your node

sudo service persistence stop

Reset your node. WARNING: This will erase your node database. If you are already running validator, be sure you backed up your priv_validator_key.json prior to running the the command. The command does not wipe the file. However, you should have a backup of it already in a safe location.

# On some tendermint chains
persistenceCore unsafe-reset-all

# On other tendermint chains
persistenceCore tendermint unsafe-reset-all --home $HOME/.persistenceCore --keep-addr-book

Since Persistence has enabled wasm and its wasm folder is outside the data folder, our snapshot also includes a wasm folder. Notice that we have taken out the cache sub-folder from the snapshot to ensure the wasm folder is compatible for all CPUs.

To sure that you have a clean start, please delete your wasm folder manually because unsafe-reset-all does not reset the wasm folder.

rm -r ~/.persistenceCore/wasm

Decompress the snapshot to your database location. You database location will be something to the effect of ~/.persistenceCore depending on your node implemention.

lz4 -c -d persistence_11674615.tar.lz4  | tar -x -C $HOME/.persistenceCore

Now double-check the ~/.persistenceCore/wasm folder to ensure that it is not empty (it is okay for Osmosis wasm folder to be empty). If it is empty, it means that our snapshot script has a bug. Please contact us via Discord Server.

If everything is good, now restart your node

sudo service persistence start

Remove downloaded snapshot to free up space

rm -v persistence_11674615.tar.lz4

Make sure that your node is running

sudo service persistence status
sudo journalctl -u persistence -f

ADVANCED ROUTE: The above solution requires you to download the compressed file, uncompressed it and then delete the original file. This requires extra storage space on your server. You can run the following combo command to stream the snapshot into your database location. For advanced users only:

curl -o - -L https://snapshots.polkachu.com/snapshots/persistence/persistence_11674615.tar.lz4 | lz4 -c -d - | tar -x -C $HOME/.persistenceCore

ALTERNATIVE ROUTE: We also have Persistence state-sync service to help you bootstrap a node.