Cockroach Labs Blog

Ideas you will meet

Databases

Schemas, indexes, query plans and choosing a store.

Consensus

Raft, Paxos and how replicas agree on one order.

Replication

Leader/follower, multi-leader and leaderless replication.

What we already know about this

Our own lessons on these ideas. Read them here before you start, so it makes sense.

• Software Design Essentials · 11 minClean and hexagonal architectureKeep business rules independent of frameworks, databases and the web: the dependency rule, layers, and ports and adapters.Read hereClose

  The problem

  In many applications, business rules are tangled with framework code, SQL and HTTP details. Then:

  • You cannot test a rule without a database and a web server.
  • Changing a framework or database means rewriting business logic.
  • The code is organised around technology, not around what the system does.

  Several architectures (hexagonal, onion, clean) solve this with the same idea: put business rules at the centre and keep details at the edges.

  The dependency rule

  100%

  Figure SDA-SDE-06-1 · Clean architecture: dependencies point inwards · © forcsguy.com

  Source code dependencies point inwards only. Entities know nothing about use cases; use cases know nothing about the web or the database. Outer layers depend on inner ones, never the reverse.

  Layer	Contains	Knows about
  Entities	Core business objects and rules that would exist whatever the technology	Nothing else
  Use cases	Application-specific actions ("place order", "cancel subscription") that coordinate entities	Entities and the ports they define
  Interface adapters	Controllers, presenters, repositories' implementations, mappers	Use cases
  Frameworks and drivers	The web framework, database, message queue, UI	Adapters

  Before: logic tangled in the controller

  100%

  After: business rules in the middle

  100%

  What changed

  • Business rules can be tested without a web server or database
  • The database or framework can be replaced at the edge
  • There are more files and layers to maintain

  Ports and adapters (hexagonal)

  The application is the hexagon. It talks to the outside only through ports: interfaces it owns. Adapters translate between the outside world and the ports.

  • Driving (inbound) ports: how the world calls the application (PlaceOrder). Adapters: REST controller, CLI, message listener, test.
  • Driven (outbound) ports: what the application needs (OrderRepository, PaymentGateway, Clock). Adapters: SQL implementation, payment provider client, fake for tests.

  // Port owned by the application
  interface OrderRepository { save(order: Order): Promise<void>; byId(id: string): Promise<Order | null> }
  interface PaymentGateway { charge(cents: number, token: string): Promise<boolean> }
  
  // Use case: pure application logic
  class PlaceOrder {
    constructor(private orders: OrderRepository, private payments: PaymentGateway) {}
    async run(input: { items: Item[]; token: string }) {
      const order = Order.create(input.items);                   // entity enforces rules
      if (!(await this.payments.charge(order.totalCents, input.token))) throw new Error("Payment failed");
      await this.orders.save(order);
      return order.id;
    }
  }
  
  // Adapters live outside, implementing the ports
  class SqlOrderRepository implements OrderRepository { /* uses a database */ }
  class StripeGateway implements PaymentGateway { /* calls the provider */ }
  
  // Composition root wires everything together at start-up
  const placeOrder = new PlaceOrder(new SqlOrderRepository(db), new StripeGateway(key));
  

  This is the dependency inversion principle from SOLID applied to a whole system.

  Benefits

  • Testable: run use cases with in-memory fakes, fast and without infrastructure.
  • Replaceable details: change the database, queue or web framework by writing an adapter.
  • Focus on the domain: the central code reads like the business process.
  • Deferred decisions: you can postpone choosing a database.

  Costs and cautions

  • More structure: interfaces, mappers and wiring add code. For a small CRUD app it can be overkill.
  • Mapping between layers (domain objects, database rows, API shapes) is repetitive; do it deliberately and avoid leaking database entities into the API.
  • Do not create layers "because the diagram says so". Add a boundary where a real change or testing need exists.
  • It is about dependencies, not folder names: a good folder layout helps but does not guarantee the rule is followed. Enforce it with tests or lint rules.

  Related ideas

  • Domain-driven design supplies the language for the centre of the hexagon (next lesson).
  • Microservices are a deployment choice; each service can use this internal structure.
  • Composition root: one place, at start-up, where concrete classes are chosen and injected.

  Real-world reading

  Related reading

  • Cockroach Labs BlogCockroach Labs engineers · Cockroach LabsBlogs
  • Fivetran BlogFivetran engineers · FivetranBlogs
  • ClickHouse BlogClickHouse engineers · ClickHouseBlogs
  • InfluxData BlogInfluxData engineers · InfluxDataBlogs

  Show 6 moreShow fewer

  • Elastic BlogElastic engineers · ElasticBlogs
  • AWS Database BlogAWS database teams · Amazon Web ServicesBlogs
  • How SQLite WorksSQLite team · SQLiteDocumentation
  • AWS Big Data BlogAmazon Web Services engineers · Amazon Web ServicesBlogs
  • Viget Engineering BlogViget engineers · VigetBlogs
  • Meituan Blog (Chinese)Meituan engineers · MeituanBlogs

  Open this lesson with its quiz and exercise

• Databases Deep Dive · 12 minReplication topologies and multi-region databasesSingle-leader, multi-leader and leaderless replication, how they handle conflicts and lag, and what changes across regions.Read hereClose

  Why replicate

  Replication keeps copies of data on several machines for three reasons: availability (survive a node failure), read scaling (serve reads from replicas) and latency (put data near users). The hard part is keeping copies consistent while writes continue. (Basics are in the replication and sharding lesson of the foundations course; this lesson compares the topologies.)

  Single-leader (primary and replicas)

  One node, the leader (primary or master), accepts writes and ships its changes to followers (replicas), which serve reads.

  100%

  Figure SDA-DDV-04-1 · Single-leader replication · © forcsguy.com

  • Synchronous replication waits for a follower before acknowledging: safer, slower, and a stalled follower can block writes. Asynchronous is faster but a failover can lose the latest writes. A common compromise: one synchronous follower, others asynchronous.
  • Failover must detect failure, choose the most up-to-date replica, redirect clients and ensure the old leader does not come back believing it still leads (split brain).
  • Replication lag causes stale reads. Techniques: read-your-writes (route a user's reads to the leader shortly after they write), monotonic reads (keep a user on one replica), and checking replica position.

  Multi-leader

  Several nodes accept writes (often one per region) and replicate to each other. Benefits: local write latency and continued operation if a region is cut off. Cost: write conflicts.

  Conflict handling options:

  • Avoid them: route all writes for one record to one leader (for example by user or region).
  • Last write wins by timestamp: simple, but silently drops one of two concurrent updates and depends on clocks.
  • Merge values (union of sets, counters) using mergeable data types (CRDTs; see the collaborative editing lesson).
  • Custom resolution in application code, or ask the user.

  Multi-leader also appears in offline-capable apps (each device is a leader) and collaborative editing.

  Before: single leader

  100%

  After: multi-leader across regions

  100%

  What changed

  • Users in each region write to a nearby database
  • Concurrent edits to the same row need conflict handling
  • More moving parts to run and monitor

  Leaderless (Dynamo-style)

  Any replica accepts writes; clients write to several replicas and read from several, using quorums (W + R > N) to get fresh data (see the key-value store lesson). Background anti-entropy and read repair fix stale copies; versions or vector clocks detect conflicts. It favours availability and is typical of Cassandra and DynamoDB-style stores.

  100%

  Figure SDA-DDV-04-2 · Choosing a replication model · © forcsguy.com

  Multi-region designs

  • Active-passive: one region serves writes; another is a warm standby. Simple; failover loses a little data (the replication lag) and takes minutes.
  • Active-active: several regions serve traffic. Requires either partitioning data by home region (each record has one home, cross-region reads are replicas) or true multi-leader conflict handling.
  • Global databases use consensus across regions (see the consensus lesson) for strong consistency at the cost of cross-region write latency; some offer tunable consistency per query.
  • Latency numbers matter: a cross-ocean round trip is around 100 ms or more, so synchronous cross-region replication adds that to every write.
  • Data residency laws may require keeping certain users' data in a region.

  Partitioning and replication together

  Real systems shard data into partitions and replicate each partition (a leader per partition spreads write load across nodes). Rebalancing, hot partitions and cross-partition transactions are the new problems (see sharding and the consistent-hashing lesson).

  Master/slave naming

  You will see "master/slave", "primary/replica" and "leader/follower" for the same single-leader idea; most current documentation prefers the last two.

  Checklist

  • What happens to writes if the leader or a region is lost, and how long?
  • How much data can we lose (RPO) and how long can we be down (RTO)?
  • How do we handle stale reads from replicas?
  • Who resolves conflicts, and how?
  • Have we practised failover?

  Real-world reading

  Related reading

  • Active-Active for Multi-Regional ResiliencyNetflix Engineering · NetflixCase Studies
  • Designing Data-Intensive ApplicationsMartin Kleppmann · O'ReillyBooks
  • Read Consistency with Database Replicas at ShopifyShopify engineers · ShopifyCase Studies
  • Uber Schemaless: A Scalable MySQL DatastoreUber Engineering · UberCase Studies

  Show 6 moreShow fewer

  • Mitigating MySQL Replication Lag and Reducing Read Load at GithubGithub engineers · GithubCase Studies
  • Cassandra architecture overviewApache Cassandra · Apache Software FoundationDocumentation
  • How Discord Stores Trillions of MessagesBo Ingram · DiscordBlogs
  • Spanner: Google's Globally Distributed Database (paper)Corbett et al. · GooglePapers
  • Scaling Multitenant Architecture Across Multiple Data Centres at Shopify - Weingarten, Engineering Lead at ShopifyShopifyYouTube
  • MIT 6.824: Distributed SystemsRobert Morris · MITCourses

  Open this lesson with its quiz and exercise

• Advanced Distributed Systems · 12 minConsensus: Raft and PaxosHow a group of machines agrees on one order of operations even when some crash or messages are delayed, and why a majority is the magic number.Read hereClose

  The problem consensus solves

  Several machines must agree on one sequence of operations (who is the leader, what is the next log entry, who owns a lock) even though machines crash and messages are delayed or lost. Without a protocol, replicas drift apart and two nodes may both believe they are the leader.

  Systems that rely on consensus include coordination services (etcd, ZooKeeper), many databases for their metadata or their data, and Kubernetes (through etcd).

  Why a majority

  A decision is valid when a majority (quorum) of nodes accepts it. Any two majorities share at least one node, so a later decision always meets someone who remembers the earlier one. A cluster of n nodes tolerates floor((n-1)/2) failures: 3 nodes tolerate 1, 5 nodes tolerate 2. That is why clusters have an odd number of nodes; a 4th node adds cost but no extra fault tolerance over 3.

  Raft in three parts

  Raft splits the problem so it can be understood:

  100%

  Figure SDA-ADS-01-1 · Raft roles and elections · © forcsguy.com

  1. Leader election. Time is divided into numbered terms. A node that hears nothing becomes a candidate and requests votes. A node votes at most once per term, and only for a candidate whose log is at least as up to date as its own. Randomised timeouts make split votes rare.
  2. Log replication. The leader appends a client command to its log and sends it to followers. When a majority has stored the entry, it is committed; the leader applies it and replies to the client. Followers apply committed entries in order, so every replica runs the same sequence.
  3. Safety. The vote rule guarantees that any new leader already has every committed entry, so committed entries are never lost or reordered.

  100%

  Figure SDA-ADS-01-2 · Replicating one write · © forcsguy.com

  Paxos, briefly

  Paxos reaches the same guarantee with two phases (prepare/promise, then accept/accepted) over a majority. It is famous for being hard to implement and explain, and for being the basis of many systems. Multi-Paxos repeats it for a log of decisions. Raft is essentially a clearer arrangement of the same ideas with a strong leader, which is why most new systems choose it.

  What consensus costs and what it does not give

  • Latency: every committed write waits for a majority round trip, so cross-region clusters are slow. Keep consensus groups within a region or use few, small groups for metadata.
  • Throughput: one leader handles all writes in a group. Large databases run many small groups (one per data range) to spread load.
  • Availability under partition: the minority side cannot make progress. This is the CAP trade-off made concrete: consensus systems are consistent first.
  • It does not remove all problems: slow nodes, clock assumptions, and reconfiguration (adding or removing nodes safely) need care.

  Where you will meet it

  Leader election for databases and queues, distributed locks and configuration stores, metadata for storage systems, and cluster state in Kubernetes. When a design needs "exactly one leader" or "all replicas agree on this order", you are describing consensus. See also the distributed-locks lesson in the core-concepts course.

  Real-world reading

  Related reading

  • In Search of an Understandable Consensus Algorithm (Raft paper)Ongaro and Ousterhout · Stanford UniversityPapers
  • The Raft Consensus Algorithm (site and visualisation)Diego Ongaro and community · raft.github.ioDocumentation
  • ZippyDB: Facebook's key-value storeMeta Engineering · MetaCase Studies
  • How Ringpop from Uber Engineering Helps Distribute Your ApplicationUber Engineering · UberCase Studies

  Show 6 moreShow fewer

  • etcd Raft libraryetcd · CNCF / etcdGitHub
  • Google SRE Book: Managing Critical State (Distributed Consensus)Google SRE team · GoogleBlogs
  • Paxos Made Simple (paper)Leslie Lamport · Microsoft ResearchPapers
  • HashiCorp RaftHashiCorp · GitHubGitHub
  • Distributed Systems (4th edition)Maarten van Steen and Andrew Tanenbaum · distributed-systems.netBooks
  • Scaling Hadoop YARN cluster beyond 10,000 nodes at LinkedInLinkedIn engineers · LinkedInCase Studies

  Open this lesson with its quiz and exercise