Replication
Replication is the creation or maintenance of multiple copies of data, state, or functionality across separate locations or components.
Kafka for Architects
Data replication is the process of copying data between systems or storage locations to maintain corresponding datasets.
- How it is used
- Uses replication to integrate external systems, keep partition data available after broker failure, and support disaster recovery.
- How deep it goes
- Explains connector flows, partition copies, and cross-site recovery trade-offs; defers leader–follower synchronization internals to an external resource.
- Examples
- Illustrates address synchronization, PostgreSQL and MongoDB sources feeding MS SQL and S3, and stretched versus active-passive clusters.
- What you can do after
- Operate/configure it by choosing connector flows and replication factors, then evaluating schema failures, restore needs, and cross-site recovery trade-offs.
Microsoft Azure in Action
Replication is the creation and maintenance of multiple copies of data or system state across storage locations or components.
- How it is used
- Uses replication to keep Azure storage and databases available across failures and, where useful, closer to readers.
- How deep it goes
- Explains service-specific replica locations, failover choices, prerequisites, CLI setup, and costs without detailing replication protocols.
- Examples
- Uses replicated digital-book blobs, a cross-region SQL ordering database, and a two-region Cosmos DB account with failover priorities.
- What you can do after
- Operate/configure it with Azure CLI storage policies, SQL replicas or failover groups, and Cosmos DB region settings.
Architecting an Apache Iceberg Lakehouse
Data replication is the process of maintaining copies of data in multiple storage systems or locations.
- How it is used
- Supports durability, high availability, and disaster recovery when selecting storage and operating Iceberg lakehouses.
- How deep it goes
- Explains storage-model tradeoffs and Iceberg recovery workflows, including metadata rewriting, but stops short of vendor-specific configuration.
- Examples
- Uses HDFS three-way replication, cloud cross-region replication, metadata path rewriting, and automated recovery DAGs.
- What you can do after
- Operate/configure replication with coordinated backups, metadata handling, catalog recovery, validation, and automated regional failover workflows.
Read how Architecting an Apache Iceberg Lakehouse explains replication →
Apache Kafka in Action
Replication is the maintenance of multiple copies of data across system components for availability and resilience, with asynchronous replication allowing temporary divergence between copies.
- How it is used
- Keeps topic partitions available across broker failures by maintaining copies on multiple running brokers.
- How deep it goes
- Explains topic-level configuration, leader–follower copying, ISRs, high watermarks, and lag-related operational trade-offs.
- Examples
- Uses a three-partition, three-replica topic, a topic-description command, and a four-replica request rejected by a three-broker cluster.
- What you can do after
- Operate/configure it with kafka-topics.sh to set a topic’s replication factor and inspect leaders, replicas, and ISRs.
Amazon Web Services in Action, Third Edition: An in-depth guide to AWS
Replication is the technique of maintaining multiple synchronized copies of data or service state.
- How it is used
- Uses replication to improve database availability and cache read capacity in AWS applications.
- How deep it goes
- Explains primary–standby failover, synchronous RDS replication, asynchronous Redis replicas, and service trade-offs without detailing replication protocols.
- Examples
- Illustrates replication with an RDS primary and standby, EBS disks, and a 22 GB Redis cache divided into six shards.
- What you can do after
- Operate/configure it by choosing RDS Multi-AZ for production databases and Redis replicas for read-heavy caches.
Latency
A replication protocol is a protocol that coordinates the creation, updating, and consistency of multiple copies of data or state.
- How it is used
- Uses replicated copies to reduce access latency while improving availability and scalability.
- How deep it goes
- Explains acknowledgment timing and replication lag, then works through a simple primary–replica protocol rather than a production-grade implementation.
- Examples
- Uses a CDN and an in-memory key–value store with Join, Snapshot, and Put messages.
- What you can do after
- Build it from scratch as a simple Rust primary–replica key–value store with snapshot-based joins and broadcast updates.
Kafka Streams in Action, Second Edition
Replication is the process of maintaining copies of records on multiple brokers to preserve data despite broker or machine failures.
- How it is used
- Protects Kafka topic-partition data and maintains availability during broker or machine failures.
- How deep it goes
- Explains leaders, followers, replication lag, ISR membership, and failover, but does not detail client behavior or every replication setting.
- Examples
- Uses a three-replica partition, leader lag during writes, and broker replication-flow diagrams.
- What you can do after
- Operate Kafka replication by selecting a replication factor and recognizing lagging followers and eligible replacement leaders.
Read how Kafka Streams in Action, Second Edition explains replication →
MongoDB 8.0 in Action, Third Edition
Replication is the maintenance of synchronized copies of database data across multiple server instances.
- How it is used
- Supports availability, disaster recovery, and distributed reads while leaving writes on the primary.
- How deep it goes
- Explains asynchronous oplog replication and diagnostic commands, while leaving Atlas replication provisioning automated.
- Examples
- Illustrates insert and deletion oplog documents plus command output showing oplog coverage and per-secondary lag.
- What you can do after
- Operate/configure it through Atlas-managed replication and inspect oplog coverage and secondary lag with MongoDB commands.
Read how MongoDB 8.0 in Action, Third Edition explains replication →
Think Distributed Systems
Replication is the maintenance of multiple copies or instances of data or functionality to reduce reliance on a single component and improve availability or performance.
- How it is used
- Improves distributed-system reliability by distributing one logical object across multiple physical resources.
- How deep it goes
- Explains logical objects, transparency, change propagation, monotonicity, leadership strategies, lag, and consistency trade-offs.
- Examples
- Uses library copies and book editions, redundant logic gates and services, replicated key-value stores, and threshold counters.
- What you can do after
- Evaluate replication designs by comparing leadership, lag, conflict, consistency, availability, and latency trade-offs.
PostgreSQL Mistakes and How to Avoid Them
Database replication is the process of copying and maintaining database data across multiple database instances.
- How it is used
- Provides standby-node redundancy and failover capability for PostgreSQL high-availability architectures.
- How deep it goes
- Explains WAL log shipping, streaming, hot and warm standby, synchronous commits, cascading, and query-conflict consequences.
- Examples
- Uses WAL shipping and streaming standbys, a cascaded HA topology, and a primary update conflicting with a standby SELECT.
- What you can do after
- Operate/configure standby replication while weighing synchronous durability against latency and diagnosing lag or standby-query conflicts.
Read how PostgreSQL Mistakes and How to Avoid Them explains replication →
Deep Learning with Python, Third Edition
The practice of copying a tensor or variable so multiple devices hold the same value.
- How it is used
- Assigns identical variable values across device-mesh axes while configuring distributed data-parallel and model-parallel computation.
- How deep it goes
- Defines replication, contrasts it with sharding, and specifies per-dimension layouts through LayoutMap.
- Examples
- Uses a (32, 64) variable, four-way sharding into (32, 16) chunks, and a Keras LayoutMap assignment.
- What you can do after
- Operate/configure variable placement through Keras LayoutMap and device-mesh layouts.
Read how Deep Learning with Python, Third Edition explains replication →
Deep Learning with R, Third Edition
Replication is the copying of the same variable value to multiple devices along a device-mesh axis.
- How it is used
- Places identical variable values across a device-mesh axis for distributed model computation.
- How deep it goes
- Defines replication, contrasts sharding, and assigns decisions per variable dimension through LayoutMap.
- Examples
- Uses a 32-by-64 variable and a two-by-four device mesh with replicated split-model copies processing sub-batches.
- What you can do after
- Apply it via Keras LayoutMap to place model variables across a device mesh.
Read how Deep Learning with R, Third Edition explains replication →
ScyllaDB in Action
Replication is the maintenance of copies of data on multiple nodes.
- How it is used
- Maintains redundant data copies across nodes to improve availability, tolerate failures, and support horizontal growth.
- How deep it goes
- Explains replica ownership, response thresholds, node loss, and recovery, but stops short of implementation mechanics.
- Examples
- Uses replicated customer-data slices and one-, two-, or three-holder responses to illustrate operational trade-offs.
- What you can do after
- Evaluate replication choices by balancing response freshness, availability, internal coordination, and tolerance for node loss.
Related concepts
- synchronous replicationcontrasts with · Synchronous replication is a replication method in which an operation is acknowledged or completed only after required replicas have confirmed the change.
- uptimeproduces · Uptime is the duration or proportion of a specified period during which a system or service is operational and available to provide its intended function.
- replicaproduces · A replica is a copy of data or state, or an equivalent running instance of a component, maintained alongside others in a replicated system.
- resilienceaddresses · Resilience is the capacity of a system or organization to withstand disruption, adapt, recover, or continue functioning acceptably.
- reliabilityproduces · Reliability is the ability of a system or process to perform its intended functions consistently and dependably under specified conditions and over time.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.