Reading this article, I understand that erasure codes are a potential solution to the data availability issue in sharding. That is, a light node can have higher certainty regarding the contents of a block when erasure codes are used.

However, I couldn't find a straightforward explanation of the concept. Can someone shed some light on it?

1 Answer 1



  • You have 2 numbers, X and Y, each in an envelope.
  • In a 3rd envelope, you store X+Y.
  • This 3rd envelope can be considered the erasure code.
  • You will always be able to compute X and Y by only losing 1 envelope.
  • The concept: to avoid losing data, you store additional data in a smart way. (In the example above, it would not help if the 3rd envelope stored X, because losing both envelopes that had X, will mean that X is lost forever.)

The source of the above example has a more interesting case. The data is X, Y, Z, W. The erasure code of X+Y+Z+W and X+2Y+4Z+8W allows the 4 pieces of data to be recovered if any 2 pieces are lost.

Related non-ELI5:



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.