D. Comer. The ubiquitous B-tree. Computing Surveys, 2(11):121--138, 1979.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....logical operation # # OR # # . Since bitwise logical operations are well supported by computer hardware, bitmap indexes are very efficient to use [21] In many data warehouse applications, bitmap indexes are better than the tree based schemes [6] 21] 30] such as the variants of B tree [9] or R tree [11] According to the performance model proposed by Jurgens and Lenz [14] the bitmap indexes are likely to be even more competitive in the future as the disk technology improves. In addition to supporting complex queries on one single table as shown in this paper, researchers have ....

Douglas Comer. The ubiquitous B-tree. Computing Surveys, 11(2):121--

....On line analytical processing (OLAP) on these datasets presents a great challenge to most software systems built around them. Currently, the predominant model of data processing is to store data as tables in relational database management systems (DBMS) and use a variant of the B tree index [8] to facilitate the searching operations. This basic strategy is inadequate for ad hoc queries on large datasets. The B tree index is e#ective if the user queries involve only one attribute or always involve similar conditions on the same set of attributes. However, many OLAP queries are ad hoc in ....

D. Comer. The Ubiquitous B-Tree. Computing Surveys, 11(2):121--137, 1979.

.... distributing the pieces of the B tree data structure. In this report we explore the use and control of replication of parts of a distributed data structure to create efficient distributed B trees. The reader unfamiliar with the basics of B trees is referred to Comer s excellent summary [Com79] In brief, the B tree formalizes in a data structure and algorithm the technique one might use in looking up a telephone number in a telephone directory, shown graphically in figure 1 1. Begin at a page somewhere near the middle of the directory; if the sought after name is alphabetically ....

D. Comer. The Ubiquitous B-tree. Computing Surveys, 11(2):121--137, 1979.

....in which the meaning of the program is built from the outset by means of features such as function names and code comments. These clues act as advance organisers , short expository notes which provide the general concepts and ideas which can be used as an aid in assigning meaning to the code [67]. The code section in Figure 2 was deliberately presented earlier without its function name. It is presented again for comparison in Figure 4 with the name and a code comment acting as an advance organiser. Increment decrement the reference count for an object static int incRefCeunt( censt ....

"The Psychology of How Novices Learn Computer Programming", Richard Mayer, Computing Surveys, Vol.13, No.1 (March 1981), p.121.

....this approach is that no one (except perhaps for the odd student in an introductory programming course) ever writes code this way. Anyone who knows how to program will never generate a program in this manner because they can recognise the problem and pull a working solution from existing knowledge [129]. This style of program creation represents a completely unnatural way of working with code, a problem which isn t helping the adoption of formal methods by programmers (the way in which code creation actually works is examined in some detail in the next chapter) This general malaise in the use ....

"The Psychological Study of Programming", B.Sheil, Computing Surveys, Vol.13, No.1 (March 1981), p.101.

....size and the communicational and computational costs are similar for all certificates stored in the tree. The operations insertion, deletion and searching are very efficient in a B tree, because they can be done in O(log #m 2# n) where n is the number of records (certificates) in the tree [13]. B trees can be optimized for search, insertion, deletion and also certification path length by choosing the parameter m properly. Easy and efficient search and sequential access to the certificates. For certificate management purposes we extend the B tree to a Merkle ....

D. Comer: The Ubiquitous B-Tree, Computing Surveys, Vol. 11., No 2., pp. 121-137, ACM, June 1979

....locked until rebalancing is complete. A state is looked up in D for each arc of R (Line 13 in Figure 3) which will generate a lot of contention if no special precautions are taken. We found an efficient way to maintain the balance of the tree by allowing concurrent access through the use of B ([5]) these are by definition automatically balanced, whereby the part of the tree that is affected by rebalancing is restricted in a suitable way. Synchronization Schemes on B trees. AB treenodemaycontain more than one searchkey which is a GSPN marking in this context. The B is said to ....

....lock of its parent (Step 3) Key 17 can then be inserted appropriately in Step 4 without the danger of back propagation. Using efficient storage methods, the organization of the data is similar to that of binary trees, whereby B states consume at most one more byte per state than binary trees [5]. Synchronization on the Stack. The shared stack, which stores the as yet unprocessed markings, is the pool from which the threads get their work. In this sense the shared stack implicitly does the load balancing and therefore cannot be omitted. Since it has to be accessed in mutual exclusion, ....

D. Comer. The ubiquitous B-tree. Computing Surveys, 11(2):121--137, 1979.

....P restricted to the items in region R only. For d = 1, the recursive nature of order decomposable problems gives an immediate tree structure, and each of the above operations can be simply implemented with O(f(N) log 2 N) time (with O(f(N) time per tree node) by using a 2 3 tree [2] or a B tree [6, 12]. However, the multidimensional version of this problem (d 1) is much more complicated: indeed, maintaining d 1 total orders on the same set S, while splitting or concatenating each order independently of the others, makes things highly non trivial due to the interplay among di erent orders. ....

....memory is discussed in [3] For the sake of presentation, we will assume here that the number of disks is D = 1; however, our results smoothly extend to D 1 parallel disks by using the disk striping technique [37] with a block size of B 0 = BD and D 0 = 1. In the simple case d = 1, B trees [6, 12] are very popular data structures that can be successfully employed in decomposable search problems analogously to concatenable 2 3 trees [2] For higher dimensions (d 1) no provably good external memory data structures for splitting and concatenating along any dimension were previously known ....

[Article contains additional citation context not shown here]

D. Comer. The ubiquitous B-Tree. Computing Surveys 11 (1979), 121-137.

....[3] our techniques can be adapted for these other models. Others have extended text indexes and multidimensional indexes to deal with structured data [20] our structural encoding is new, and we deal with all of the structure in one index. The Index Fabric is a balanced structure like a B tree [9], but unlike the B tree, scales well to large numbers of keys and is insensitive to the length or complexity of keys. Diwan et al. have examined taking general graph structures and providing balanced, disk based access [14] Our structure is optimized specifically for Patricia tries. Query B I Os ....

D. Comer. The ubiquitous B-tree. Computing Surveys 11(2): 121-137, 1979.

....at a leaf of the tree often cause reorganizations which may affect the whole tree structure. Therefore the part of the tree that is potentially affected by the reorganization has to be locked. In our solution data consistency with a high degree of parallelism is achieved by using B trees [9] as search data structure S. In B trees a parameterizable number of markings is grouped together to one B tree node which is the entity that is locked thus reducing the number of locking operations. Rebalancing is done by splitting B tree nodes when they are full. The height of the tree is only ....

D. Comer. The ubiquitous B-tree. Computing Surveys, 11(2):121--137, 1979.

....evaluation of complex analytical queries. A very successful means to speed up query processing is the exploitation of index structures. Several index structures have been applied to data warehouse management systems (for an overview see [2, 17] Among them are traditional index structures [1, 3, 6], bitmaps [15] and R tree like structures [9] Permission to copy without fee all or part of this material is granted provided that the copies are not made or distributed for direct commercial advantage, the VLDB copyright notice and the title of the publication and its date appear, and notice ....

....TPC D data) Our goal was to design an index structure that allows for efficient support of complex queries against high volumes of data as exemplified by the TPC D benchmark. The main problem encountered is that some queries refuse the application of a (traditional) index structure (like B Trees [1, 3] and Extendible Hashing [6] due to efficiency reasons. A typical situation is, when e.g. more than one tenth of a relation qualifies for a selection predicate. Then the only effect of using an index is to turn sequential I O into random I O (in the presence of a non clustered index) Even worse, ....

D. Comer. The ubiquitous B-tree. Computing Surveys, 11(2):121--137, 1979.

....occupy all available space. 4.2 The libba Block Allocation Library Not all applications that benefit from the ability to manage physical disk space are amenable to implementation in the context of TPIE. In particular, applications that use external memory dynamic data structures, such B trees [Com79] grid files [NHS84] or R trees [Gut84] and their variants [BKSS90, SRF87] need to be able to allocate and deallocate space a block at a time. These data structures are widely used in a variety of database systems, thus there is ample motivation to support them. Without zero( it is possible ....

D. Comer. The ubiquitous B-Tree. Computing Surveys, 11(2):121--137, 1979.

....the index(es) that minimize E. Since all 2 dimensional approximate queries have the same rectangle side ( 1 vmax , 1 vmin ) Figure 4) the rectangle range search is equivalent to a simple range search on the b coordinate axis. Thus each of the c observation indices can simply be a B tree [13]. To process a general query interval [y 1q ; y 2q ] we consider two cases depending on whether the query interval covers a subterrain: i) y 2q Gamma y 1q ymax c : then it can be easily shown that area E is bounded by: E 1 2 ( v max Gamma v min vmin Theta vmax ) 2 ( ymax c ) ....

D. Comer. The Ubiquitous B-Tree. Computing Surveys, 11(2):121--137, June 1979.

....increasing. The size of any present day database is usually in the range of gigabytes. The massive size of a database is one of the main problems in designing efficient data structures and algorithms to process them efficiently. Most frequently used data structures in databases, are B trees [1] or its variations as main, and a combination of dynamic and static hashing [2, 3, 4, 5] as auxiliary data structures. B tree and its variations are well suited for disk based systems, but they are unable to grow and shrink efficiently dynamically. A dynamic data structure called Extendible ....

Comer, D., The Ubiquitous B-Tree. Computing Survey, Vol. 11, No. 2, June 1979.

....the bad worst case behaviour of these trees, various strategies have been developed to maintain them in balance. Global balancing algorithms periodically rebuild the whole tree [3, 9, 15] Balanced trees, e.g. AVL trees and red black trees, perform rebalancing coupled with each update operation [1, 5, 6, 8, 11, 13]. These data structures and algorithms were developed and analyzed assuming the Random Access Memory (RAM) model [2] This was a reasonable assumption when memory was slow and processors even slower. However, a long term trend has been that processor speed has doubled every two years and memory ....

D. Comer, The Ubiquitous B-tree. Computing Surveys 11 (1979), pp. 121--137.

No context found.

D. Comer. The ubiquitous B-tree. Computing Surveys, 2(11):121--138, 1979.

No context found.

D. Comer. The Ubiquitous B-Tree. Computing Surveys, 11(2):121--137, 1979.

No context found.

D. Comer. The Ubiquitous B-Tree. Computing Surveys, 11(2):121--137, 1979.

No context found.

D. Comer. The Ubiquitous B-tree. Computing Surveys, 11(2):121--138, 1979.

No context found.

D. Comer. The ubiquitous B-tree. Computing Surveys, 11(2):121--137, 1979.

No context found.

D. Comer. The Ubiquitous B-Tree. Computing Surveys, 11(2):121--137, June 1979.

No context found.

D. Comer. The Ubiquitous B-Tree. Computing Surveys, 11(2):121--137, 1979.

No context found.

D. Comer. The ubiquitous b-tree. Computing Surveys, 11(2):121-137, 1979.

No context found.

D. Comer. The Ubiquitous B-Tree. Computing Surveys, 11(2):121--137, June 1979.

No context found.

D. Comer. The ubiquitous b-tree. Computing Surveys, 11(2):121--137, June 1979.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC