Fast synchronization

At the time of writing this book, the blockchain size is around 30 GB. Downloading it may take several hours or days if you have a slow Internet connection. Ethereum implements a fast synchronization algorithm, which can download the blockchain faster.

Fast synchronization doesn't download the entire blocks; instead, it only downloads the block headers, transactions receipts, and the recent state database. So, we don't have to download and replay all transactions. To check blockchain integrity, the algorithm downloads a full block after every defined number of blocks. To learn more about fast synchronization algorithm, visit https://github.com/ethereum/go-ethereum/pull/1889.

To use fast sync while downloading the blockchain, you need to use the --fast flag while running geth.

Due to security reasons, fast sync will only run during an initial sync (that is, when the node's own blockchain is empty). After a node manages to successfully sync with the network, fast sync is forever disabled. As an additional safety feature, if a fast sync fails close to or after the random pivot point, it is disabled as a safety precaution, and the node reverts to full, block-processing-based synchronization.