Dealing with Storage Issues on the Execution Client

It is currently recommended to use a minimum 1TB hard disk.

Kudos to angyts for this contribution.

After running the execution client for a while, you will notice that it will start to fill up the hard disk. The following steps might be helpful for you.

Firstly make sure you have a fallback execution client: see 8.11 Strategy 2.

Manually Pruning Geth

Since Geth 1.10x version, the blockchain data can be regularly pruned to reduce it's size.

Reference: https://gist.github.com/yorickdowne/3323759b4cbf2022e191ab058a4276b2

You will need to upgrade Geth to at least 1.10x. Other prerequisites are a fully synced execution engine and that a snapshot has been created.

Stop your execution engine

sudo systemctl stop eth1

Prune the blockchain data

geth --datadir ~/.ethereum snapshot prune-state

🔥 Geth pruning Caveats:

• Pruning can take a few hours or longer (typically 2 to 10 hours is common) depending on your node's disk performance.

• There are three stages to pruning: iterating state snapshot, pruning state data and compacting database.

• "Compacting database" will stop updating status and appear hung. Do not interrupt or restart this process. Typically after an hour, pruning status messages will reappear.

Restart execution engine

sudo systemctl start eth1

Adding new hard disks and changing the data directory