Block Structure


A block is the primary data structure by which information about the state of the CasperLabs system is communicated between nodes of the network. We briefly describe here the format of this data structure.

Protobuf definition

Messages between nodes are communicated using Google’s protocol buffers. The complete definition of a block in this format can be found on GitHub ; the description here is only meant to provide an overview of the block format; the protobuf definition is authoritative.

Data fields

A block consists of the following:

  • a block_hash

  • a header

  • a body

  • a signature

Each of these are detailed in the subsequent sections.


The block_hash is the blake2b256 hash of the header (serialized according to the protobuf specification).


The block body contains a list of the deploys processed as part of this block. A processed deploy contains the following information:

  • a copy of the deploy message which was executed (see Execution Semantics for more information about deploys and how they are executed)

  • the amount of gas spent during its execution

  • a flag indicating whether the deploy encountered an error

  • a string for an error message (if applicable)


The block signature cryptographically proves the block was created by the validator who’s public key is contained in the header. The signature is created using a specified algorithm (currently only Ed25519 is supported), and is signed over the block_hash so that it is unique to that block.