Partition tolerance

As explained in PostgreSQL Mistakes and How to Avoid Them

Partition tolerance is the ability of a distributed system to continue operating despite dropped or delayed network messages.

Partition tolerance means continuing despite dropped or delayed network messages. It forces distributed systems to trade consistency against availability during partitions, while PACELC adds latency versus consistency when networks are healthy. section 5.5

The multi-master replication trap

Bidirectional PostgreSQL logical replication can look healthy: both nodes display the same tickets. But independent identity generation can assign an existing ID, producing a duplicate-key error. Renaming opened_at on one node then causes the replication worker to stop because the replicated column is missing. High availability therefore requires explicit compromises around consistency, schema coordination, and operational complexity. section 5.5

Figure 5.2 Logical replication in both directions

Use the bidirectional replication diagram to trace how independent PostgreSQL instances exchange changes and where partition-related conflicts can arise. section 5.5

How PostgreSQL Mistakes and How to Avoid Them covers it

How deep it goes

Defines CAP and PACELC, then demonstrates replication conflicts and schema failure without developing theorem proofs or recovery procedures.

Examples

Uses bidirectional PostgreSQL ticket replication, duplicate identity values, and a replicated-column rename that stops the worker.

What you can do after

Evaluate/understand distributed-system tradeoffs before selecting PostgreSQL replication and multi-master designs.

How it connects

Connects consistency, availability, CAP theorem, PACELC, logical replication, and multi-master replication.

liveBookmanning.com

Read PostgreSQL Mistakes and How to Avoid Them on liveBook Get the book at manning.com

Also explained in

• 

  liveBookmanning.com
  ScyllaDB in ActionNetwork partition tolerance is the ability of a distributed system to continue operating despite disconnected nodes or network partitions.

Compare how 2 books explain partition tolerance →

Related concepts

• CAP theoremused by · The CAP theorem is a distributed-systems impossibility result stating that consistency, availability, and partition tolerance cannot all be guaranteed during a network partition.
• network partitionuses · A network partition is a failure condition in which disrupted connectivity prevents groups of otherwise operating components in a distributed system from communicating reliably.
• partitionaddresses · A partition is a distinct subdivision of a larger data, storage, or work unit that can be stored, managed, or processed separately.
• fault tolerancekind of · Fault tolerance is the ability of a system to continue providing correct or acceptable service despite failures in its components or operating environment.

This explainer was generated by Manning's knowledge graph from the text of PostgreSQL Mistakes and How to Avoid Them. The section links point to the parts of the book it draws on.