How does multichain deals with some unreliability problems that exist in LevelDB?

+2 votes
LevelDB has been categorized as unreliable due to it is prone to errors based on studies on previous versions. Apparently it can cause data corruption after systems crash or power failures. How does multichain deals with this type of bug when using LevelDB to store streams?
asked Aug 16, 2018 by Yosh

1 Answer

+1 vote
Data loss problems on hard shutdown / power failure are expected and MultiChain has a lot of logic that runs on startup to recover this kind of lost data, by rolling back the node's internal database and replaying events on the blockchain. But we do not know of data *corruption* problems in LevelDB due to hard shutdowns and have not seen evidence of them. Can you please post a link to these studies you mention?
answered Aug 16, 2018 by MultiChain