distributed_computing:transactions

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Next revision		Previous revision
distributed_computing:transactions [2019/05/17 23:13] – created phreazerdistributed_computing:transactions [2019/05/17 23:22] (current) – [Scenario] phreazer
Line 11: Line 11:
  
 "In presence of a network partition, one has to choose between **consistency** and **availability**". "In presence of a network partition, one has to choose between **consistency** and **availability**".
 +
 +==== Consistency ====
 +  * Strong consistency
 +  * Timeline consistency: Typically async master-slave replication.
 +  * Eventual consistency: Temporarily inconsistent
 +
 +====== Scenario ======
 +
 +Scaling out NOSQL across regions/public clouds. Partitioned db across clusters. How to ensure that transaction is commited on all nodes.
 +
 +Ensuring atomicity: Multi-phase commit, pessimistic read-write locking
 +But: Network partitions
  
  
  • distributed_computing/transactions.1558127603.txt.gz
  • Last modified: 2019/05/17 23:13
  • by phreazer