Summary of Disaster Recovery Plan for Crescendo Mainnet launch
When an issue is discovered on production network (TN or MN), we always first evaluate fix by deploying a new software version (roll-forward approach to fixing an issue)
Situations we might need to use more disruptive methods to minimize the value at risk:
🗝️ Exploit of a vulnerability leading to resource duplication or unauthorized account manipulation.
📀 Bug in a storage layer causing corruption of the on-chain data which is hard to revert. (hard to revert means that the effort to revert to uncorrupted state would take days, not hours).
📜 Security / usability issue that requires Cadence Language breaking change to be fixed.
💔 Broken dapp functionality as a result of a bug in the state migration.
Methods we will consider to reduce impact, fix and recover the network: