Skip to main content

CAP theorem demystified


In a web world, everything is available over the internet. While designing a distributed application for the client, we often face this dilemma of choosing between consistency and availability.  There is no further argument if we don't choose to have a partition tolerant system, that's single node application and it will adhere to consistency and availability. So we will be discussing Partition tolerant systems for a distributed environment.  Lets first understand CAP individual elements;

C (Consistency): 
In a distributed environment of n nodes, a change in one node should be instantly reflected all other nodes. So any data change/state should be reflected all client irrespective of whichever node they access the data. 


A (Availability):

All the non-failing node should be ready to serve any request within a reasonable time. This applies to every node present in the system. 

P (Partition Tolerence):

Distributed systems are meant to build partition tolerant systems. Any node creation and removal should be gracefully handled. 


A distributed system need to partition tolerant, So P axis will be fixed. As there is a contradiction between Consistency and availability, we have to fine-tune them for business cases. 

Example: I have 3 servers with user data in place. Now All three have the same data, so any read request will be having consistent data for any read request to any node.  Now if user A has modified his details and call has been made to server S1 at T0. So data is updated at S1 at T0 but will be propagated to S2 and S3 post T4 (Assume). So either S2 and S3 will choose to decline any request till T4 to maintain consistency and compromise availability or will continue to serve request but with an outdated data. 

And that's gist of CAP theorem. 

Comments

Post a Comment

Popular posts from this blog

[PUZZLE] ELEPHANT AND BANANA PUZZLE

A merchant has bough some 3000 banana from market and want to bring them to his home which is 1000 km far from Market. He have a elephant which can carry maximum of 1000/- banana at time. That could have been easy but this elephant is very shrewd and he charges 1 banana for every one kilometer he travel. Now we have to maximise number of banana which should reach to home. Solution: At present we are at 0th km with 3000 banana and elephant. Lets see if elephant have to shift all the 3000 banana and elephant by 1 km. For 1 km shift: Take first 1000 banana and drop at 1st km. 2 banana consumed. One banana each for to and fro. Second  1000 banana and drop at 1st km. 2 banana consumed. Third 1000 banana and reach at 1st km. 1 banana consumed. So all in all total 5 banana will be consumed if we shift a kilometer. But that's not all, our total banana number are also reducing so we may have to reduce the number of turns too. And this will happen once we have reduced the b
Post a Comment
Read more

[JAVA] Evil Null

Elvis operator could have been a good use here. But unfortunately have been decline till now. Its used to say "?." Operate only if not null. Class A {  public B getB() {    return new B(); } } public void test( A a) {     if ( a != null  )         return a.getB();  } Above method would be replaced with public void test ( A a) {      return a?.getB();  } Unfortunately its still some time we can see some think like that. So till that we have to live with two choices: 1. Check for null. 2. Use the Null Object Pattern. So we all know that how to check for null objects and believe me i had real long chain of checking null. Second way can be useful but if this chain is pretty long its can be tiresome to have null object for all the hierarchy Null Object means there will be definition of class A (mostly derived from same interface.) and will have dummy methods which will nullify any call to these methods.
Post a Comment
Read more

Two numbers with minimum difference

Find the two numbers whose difference is minimum among the set of numbers. For example the sequence is 5, 13, 7, 0, 10, 20, 1, 15, 4, 19 The algorithm should return min diff = 20-19 = 1. Make an efficient algorithm, yeah best could be O(n). Not sure if i could use DP here, will post a solution tommorow. Let me know if you find a answer. Algorithm:MiniMumDiffN Get the size of array , say n. Get the range of array, range. Bucket Sort the above array. Search the above sorted array for minimum difference. This will be the minimum difference numbers. Now question comes where the range of above given array is very big. Algorithm:DivideArray Take an random element. Search for its position in array. Return the index of element. Algorithm:MiniMumDiff if array_range > MAXRANGE int mid = divideArray (arr,start,end); d1,max1 = MiniMumDiff( arr,start,mid-1); d2,max2 = MiniMumDIff(arr,mid+1,end); d3 = difference max1 and arr[mid] d4 = difference max2 and arr[mid] return minimum( d1,d2
1 comment
Read more