Cap and base theory

1. CAP theory
   2000 year 7 month , University of California, Berkeley Eric Brewer Professor at ACM PODC Proposed at the meeting CAP guess .2 After year , MIT's Seth Gilbert and Nancy Lynch It is proved theoretically CAP. after ,CAP Theory has become an accepted theorem in the field of distributed computing .

CAP The theory is ： A distributed system can only satisfy the consistency at most （Consistency）、 Usability （Availability） And partition fault tolerance （Partition tolerance） Two of these three .

1.1 Uniformity （Consistency）
Consistency means “all nodes see the same data at the same time”, That is, after the update operation succeeds and returns to the client , All nodes have the same data at the same time .

1.2 Usability （Availability）
Usability means “Reads and writes always succeed”, That is, the service is always available , And it's normal response time .

1.3 Partition tolerance （Partition tolerance）
Partition fault tolerance means “the system continues to operate despite arbitrary message loss or failure of part of the system”, That is, when the distributed system encounters a node or network partition failure , Still able to provide external services to meet the consistency and availability .

2. CAP Balance
   adopt CAP theory , We know that consistency cannot be met at the same time 、 Availability and partition fault tolerance , Which one should I give up ？

For most large Internet application scenarios , There are many hosts 、 Deployment is decentralized , And now the scale of the cluster is growing , So node failure 、 Network failure is the norm , And make sure the service availability reaches N individual 9, Guarantee P and A, Abandon C（ Second, to ensure the final consistency ）. Although some places can affect the customer experience , But it didn't reach the level that caused the user process .

There can't be a little concession when it comes to money ,C Must ensure . In case of network failure, it's better to stop service , This is a guarantee CA, Abandon P. It seems that in recent years, the domestic banking industry has been on the rise 10 There was an accident , But it doesn't matter , There's not much reporting , The masses know little . There is also a guarantee CP, Abandon A. For example, if there is a network fault, it is read-only and not written .

Which is better, which is less , No conclusion , It can only be decided according to the scene , What suits is the best .

3. BASE theory
   eBay Architect Dan Pritchett From the practice summary of large-scale distributed system , stay ACM Put forward in an article published on BASE theory ,BASE The theory is right CAP The extension of theory , The core idea is that even if strong consistency cannot be achieved （Strong Consistency,CAP Consistency is strong consistency ）, But the application can achieve final consistency in a suitable way （Eventual Consitency）.

BASE Basically available （Basically Available）、 Soft state （ Soft State）、 Final consistency （ Eventual Consistency）.

3.1 Basic available （Basically Available）
Basic availability refers to the failure of distributed system , Partial loss of availability is allowed , That is to ensure that the core is available .
When e-commerce is booming , In response to the surge in traffic , Some users may be directed to the degraded page , The service layer may only provide degraded Services . This is the loss of some availability .

3.2 Soft state （ Soft State）
Soft state means that the system is allowed to have intermediate state , The intermediate state will not affect the overall availability of the system . In distributed storage, there are at least three copies of a data , The delay that allows the replica synchronization between different nodes is the embodiment of the soft state .mysql replication Asynchronous replication is also a manifestation of .

3.3 Final consistency （ Eventual Consistency）
Final consistency means that after a certain period of time, all data copies in the system , In the end, we can reach a consistent state . Weak consistency is the opposite of strong consistency , Final consistency is a special case of weak consistency .
4. ACID and BASE The difference and connection
ACID It is a common design concept of traditional database , The pursuit of a strong consistency model .BASE It supports large distributed systems , High availability is achieved by sacrificing strong consistency .

ACID and BASE Represents two completely opposite design philosophy

In the scenario of distributed system design , System components have different requirements for consistency , therefore ACID and BASE It will be used in combination with .