Question

between B trees and B+ trees,which one is suited for random and which one is suited for sequential access??

please explain

Answer 1

Both B and B+ trees are pratical implementations of sorted file organisation. I am not sure what you mean by sequential access. But as the records are sorted they can be accessed in sequence. And in B+ trees all leaf nodes are linked and they contain all index key values and record pointers. Hence they are perfect for range queries. for eg where salary between 1000000 and 200000.

For quick random access on the other hashed file organisation is best suited.

Comments

In B+ trees, if the searching is done on the element(s) that are indexed using B+ tree, B+ tree definitely gives better performance.

Else, its difficult to compare.

---

Also, hasing may give bad performance as compared to B trees. Depends upon how hashing is implemented i.e. chaining may give as good performance as B trees.

---

If the hashing algorithm used uniformly distributes the data such that minimal conflicts occurs then the elements can be accessed in O(1). And chaining is used to resolve conflicts in hashing. Dynamic and extensible hashing enables the hash bucket to grow dynamically eliminating conflict all together. So hashed file organisation is best suited for random access.

---

is it practically possible to distribute uniformly? I mean I am just asking :) . Yes, but atleast therotically, its O(1)