Fastest fixed key / db value for fixed size keys and only insert / select (no delete / update)?

Question

Given the following requirements for a persistent key / value store:

• Only fetch, insert, and complete iteration of all values ​​are required (for export)
• Missing values ​​or updating values
• The keys are always the same size.
• Code embedded in the host application

And given this usage pattern:

• Elections are random.
• Insertions and selections alternate without predictability
• Keys are random and inserted randomly.

What is the best structure / algorithm on disk given the requirements?

Can a user implementation exceed the performance of LSM (Log Structured Merge) based implementations (i.e. leveldb, rocksdb)?

?

Answer 1

, , RocksDB . RocksDB:

-, , RocksDB. RocksDB , , , . , - RocksDB, 1. , , ; 2. ; 3. . , , .

-, , IO, , , , RocksDB.

, . -, , , . , (), , , , . , .

, RocksDB , RocksDB.

RocksDB .