Skip to content

Latest commit

 

History

History
71 lines (54 loc) · 2.98 KB

syncing-your-node.md

File metadata and controls

71 lines (54 loc) · 2.98 KB
Raw

Sync your Avalon node

Once your avalon node and dependencies are setup correctly, if you run it with ./scripts/start.sh, it will start a new developement chain. Instead, you probably want to sync your node to the mainnet.

First, we need to download the genesis block (block #0) and save it as ./genesis/genesis.zip:

mkdir genesis
cd genesis
wget https://backup.d.tube/genesis.zip
cd ..

Secondly, you will need to add some default peers to use to connect to the network. Near the bottom of the scripts/start.sh script:

export PEERS=ws://34.65.228.228:6001,ws://dseed.techcoderx.com:6001
export MAX_PEERS=50

If you need a bigger peer list, come to our discord server and look at the pinned messages of the #leader-candidates channel.

You would also need to specify the path to the folder containing blocks.bson file in BLOCKS_DIR, which is a read and append only file that stores all blocks.

export BLOCKS_DIR=/path/to/blocks/dir

Then finally you want to start replaying the blocks. Unless this is a brand new node install, your avalon database will probably contain data. You will need to wipe it before starting a replay.

You can wipe the mongodb by doing mongosh avalon --eval "db.dropDatabase()" (assuming you are using the default 'avalon' db name).

Finally, to replay the blocks and transactions, you have 3 options:

1. Natural replay

This is the easiest method. Just start the node with ./scripts/start.sh and you should see your node unzipping the genesis data, and then starting to download blocks from the peers. This method can be very slow, and probably not scalable in the long term.

2. Rebuild from blocks BSON

This is the fastest method that reverifies locally all the past blocks and transactions, and therefore the current blockchain state. You will need to download the blocks.bson file into the folder specified in BLOCKS_DIR env var.

cd /path/to/blocks/dir
wget -c https://backup.d.tube/blocks.bson
cd /path/to/avalon/directory
REBUILD_STATE=1 ./scripts/start.sh

A file named blocks.index will be constructed (if not already exists) which stores the pointers to each block in blocks.bson. This may take a few minutes.

To speed up the rebuild process further, you may opt to skip block validations with REBUILD_NO_VALIDATE=1 env var.

3. Replay from database snapshot

This is the fastest method (takes <5 mins). You will download the latest hourly snapshot and import the data in your node, without any verification.

mkdir dump
cd dump
wget https://backup.d.tube/$(date -u +%H).tar.gz
tar xfvz ./*
mongorestore -d avalon ./
cd ..
rm -rf ./dump
./scripts/start.sh

Creating your own dumps for quick replays

Alternatively, if you do not want to have to trust our backups (coming from d.tube domain), you can create your own:

First, shut-down your node to avoid any new incoming data contaminating your backup. Then just run:

mongodump -d avalon -o ~/avalon-backup/

And restart your node