What is this?
RocksDB is an embeddable, persistent key-value store for fast storage, built as a C++ library and developed by Facebook's Database Engineering team.
RocksDB is an embeddable, persistent key-value store for fast storage, built as a C++ library and developed by Facebook's Database Engineering team.Is this a database server I deploy, or a library I link into my app?A library, not a server: you embed it in your application. The README describes it as the core building block you would use to build a fast key-value server.
A library, not a server: you embed it in your application. The README describes it as the core building block you would use to build a fast key-value server.How do I start using RocksDB in my own project?Start with the example usage linked in the README, then the GitHub wiki for build and usage details. As a C++ library, you compile it and link it into your application.
Start with the example usage linked in the README, then the GitHub wiki for build and usage details. As a C++ library, you compile it and link it into your application.Which parts of the API are safe to build against?Only the headers in include/ — that is the public interface. Anything else is internal and may change without warning between versions.
Only the headers in include/ — that is the public interface. Anything else is internal and may change without warning between versions.How much data can a single RocksDB database hold?Multiple terabytes in one database. Multi-threaded compactions are what make handling data at that scale practical.
Multiple terabytes in one database. Multi-threaded compactions are what make handling data at that scale practical.What storage is it built for — SSDs, spinning disks, or memory?Flash and RAM storage: the README's title is 'A Persistent Key-Value Store for Flash and RAM Storage', and it says the library is especially suited for storing data on flash drives.
Flash and RAM storage: the README's title is 'A Persistent Key-Value Store for Flash and RAM Storage', and it says the library is especially suited for storing data on flash drives.What can I tune about how it balances write, read and space costs?Its LSM (Log-Structured-Merge-Database) design offers flexible tradeoffs between Write-Amplification-Factor (WAF), Read-Amplification-Factor (RAF) and Space-Amplification-Factor (SAF), so you can bias it toward write speed, read speed or disk footprint.
Its LSM (Log-Structured-Merge-Database) design offers flexible tradeoffs between Write-Amplification-Factor (WAF), Read-Amplification-Factor (RAF) and Space-Amplification-Factor (SAF), so you can bias it toward write speed, read speed or disk footprint.How is it related to LevelDB?It is built on earlier work on LevelDB by Sanjay Ghemawat and Jeff Dean, and is now developed and maintained by Facebook's Database Engineering team.
It is built on earlier work on LevelDB by Sanjay Ghemawat and Jeff Dean, and is now developed and maintained by Facebook's Database Engineering team.What are similar projects?dragonfly, scylladb, badger, nebula, yugabyte-db, foundationdb, dicedb, KeyDB
dragonfly, scylladb, badger, nebula, yugabyte-db, foundationdb, dicedb, KeyDBWhich topics is it in?storage-engine, key-value-store, embedded-database, multithreading, scalability, nosql, storage, gplv2, high-performance, cpp, c++, database
storage-engine, key-value-store, embedded-database, multithreading, scalability, nosql, storage, gplv2, high-performance, cpp, c++, database