Topic: What advantages does using LevelDB offer? (Read 1422 times)
Thanks for all your replies, I have a clearer picture now.
The UXTO db as implemented stores unspent outputs, serialized, keyed by txid.
That's it. Want to know if a txid has unspent outputs? Look in the DB and deserialize the information.
Nothing more advanced than that is really required.
That's it. Want to know if a txid has unspent outputs? Look in the DB and deserialize the information.
Nothing more advanced than that is really required.
What are the advantages? Why was it implemented?
I've used LevelDB, H2, Firebird and PostgreSQL with my JavaBitcoin node. LevelDB wins hands-down on both performance and database size. It is ideal for a bitcoin database where you have a small number of keys (usually just the transaction or block hash) and don't need the complexity of the SQL SELECT statement. It is easy to implement a secondary index (such as block height->block) using a separate map.Ron
When you are dealing with data that is mostly *binary* in nature a SQL DB doesn't really offer much advantage over a DB that is more "object based".
Basically you end up having to do *more work* with a SQL DB than a non-SQL DB to store/retrieve the records - so if you don't need SQL then it is actually going to be *faster* to not use it.
Understand that the *simplest* way to do things will always be the *fastest*.
Basically you end up having to do *more work* with a SQL DB than a non-SQL DB to store/retrieve the records - so if you don't need SQL then it is actually going to be *faster* to not use it.
Understand that the *simplest* way to do things will always be the *fastest*.
What are the advantages? Why was it implemented?
Jump to: