You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
To be protected against the attack outlined in https://arxiv.org/pdf/1902.10244.pdf we need AuRa to require 2/3 of signatures instead of 1/2 as it was proposed by the researches in the section 9.2.2 of the article. Can we do that?
The text was updated successfully, but these errors were encountered:
If I run from scratch a few nodes with 1/2 majority and the rest of the nodes with 2/3, they fail to apply a new validator set when staking epoch changes (a fork occurs for a part of the nodes).
If I update all the nodes before changing the validator set, they all continue to work successfully. The nodes which are not updated get error and terminate.
So, to switch to supermajority successfully, all the nodes must be updated before the validator set is changed.
We must introduce supermajority activation with some spec option which would turn it on for all the nodes at the specified block simultaneously (e.g. supermajorityTransition). After that, we could retest the switching to supermajority again to ensure we can safely do this for Sokol/Core/Dai/Kovan.
To be protected against the attack outlined in https://arxiv.org/pdf/1902.10244.pdf we need AuRa to require 2/3 of signatures instead of 1/2 as it was proposed by the researches in the section
9.2.2
of the article. Can we do that?The text was updated successfully, but these errors were encountered: