Open skhoroshavin opened 4 years ago
We've onboarded a number of Stewards onto the various Sovrin networks since this issue was reported. The new nodes do catch up and reach consensus, but it takes a few hours (3 or so) for this to occur (without seeding the node with an ledger data).
I have seen the issue @skhoroshavin saw on the Indicio TestNet. Not quite sure of the circumstances, but large catchups caused issues for sure.
It looks like now Indy Node/Plenum cannot successfully catch up when node is too much behinds (for example new node joining 3-years old pool with 10000s transactions). While workaround is possible (just copy data) it would be much better if node could catch up using normal process.
More details (including logs with failed catch-up) can be found here: https://sovrin.atlassian.net/browse/SN-18
Acceptance criteria