Guides
Task-oriented walkthroughs for building on kv: transactions and isolation, choosing and tuning an engine, durability, encryption, backup and replication, running the server, and working from the command line.
Each guide is built around a job rather than an API call: keeping writes consistent under concurrency, picking the engine that fits your workload, deciding how much durability to pay for, encrypting a database at rest, getting data safely off the box, serving it over a socket, and driving it from the shell. They assume you have worked through the quick start.
Transactions and isolation
How kv runs reads and writes inside ACID transactions, what snapshot isolation guarantees, when to reach for serializable, and how to handle conflicts.
Choosing an engine
The B-tree and LSM storage cores: how each one trades read cost against write cost, when to pick which, and the tuning knobs each exposes.
Durability and recovery
How kv survives a crash through a write-ahead log, what each synchronous level guarantees, and how checkpointing and vacuum keep the file in shape.
Encryption at rest
Encrypting a kv database with AES-256-GCM, how the key is supplied and how pages are protected, and how to rotate the key without taking the database offline.
Backup and replication
Taking a consistent online backup, restoring it, and shipping the write-ahead log to keep a read replica or an archive for point-in-time recovery.
Running the server
Serving a kv database over HTTP/JSON and a binary protocol, with authentication, TLS and mTLS, rate and connection limits, and a change feed.
Working from the command line
Driving kv from the shell: reading and writing keys, scanning, moving data in and out, the interactive shell, and inspecting a database's health.