Re: B+ tree - help

From: Bob Badour <bbadour_at_pei.sympatico.ca>
Date: Wed, 27 Sep 2006 12:07:15 GMT
Message-ID: <TPtSg.40990$9u.350293_at_ursa-nb00s0.nbnet.nb.ca>


accpactec_at_hotmail.com wrote:
> Alt 2 is pretty is easy. it is basically when there is <k*, rid> such
> that k* is the data entry and the rid is the record ID pointing to the
> data from the index.
>
> Data pointer are the pointers in the case of Alt 2 is the rid. subtree
> pointers are the pointers that are being pointed from a parent to its
> child/leaf.
>
> I am not sure what you mean by "how many 'kinds' of block will you have
> in your ...".

How often do you mix sub-tree pointers and data pointers in the same block?

> Since the block size is 512, according to the specs of b+ tree, at
> least 256 byte must be filled since d < m < 2d where d is the order.
>
>
> Bob Badour wrote:
>

>>accpactec_at_hotmail.com wrote:
>>
>>
>>>Ok, I am looking at this question about b+ trees and I just don't get
>>>it. Here is the question:
>>>
>>>A given B+-tree is stored on a disk with blocks containing 512 bytes
>>>each. The indexed key, data pointer and pointer to a sub-tree occupy 8
>>>bytes, 6 bytes and 4 bytes, respectively. Assume we use Alternative (2)
>>>for data entries.
>>>
>>>a.	Compute the order of the B+-tree.
>>>b.	Calculate the minimum number of entries of data records that a
>>>2-level B+-tree (not counting the root) with the given parameters can
>>>index.
>>>
>>>I have been looking at wikipedia,
>>>http://www.semaphorecorp.com/btp/algo.html and various other sites but
>>>none of them really explain how this thing is done. Can someone please
>>>help me with this.
>>
>>What does your textbook say about Alternative 2?
>>
>>What is the difference between a data pointer and a sub-tree pointer?
>>
>>How many kinds of blocks will you have in your B+-tree index?
>>
>>For each kind of block, assume you have N keys:
>>   How many pointers will you have in the block?
>>   What is the largest N you can fit in the block?

>
>
Received on Wed Sep 27 2006 - 14:07:15 CEST

Original text of this message