A concurrent search structure

https://doi.org/10.1016/0743-7315(89)90020-8Get rights and content

Abstract

The sibling trie is a highly concurrent dynamic search structure. It supports search, update, insertion, and deletion. The sibling trie is designed to minimize “hot spots” in highly concurrent shared memory environments. The main novelty of this data structure is that searches can start from any node. As a result, many alternate routes are provided to each datum, and the bottleneck of a single root node is avoided. Search time is logarithmic (regardless of the starting point), and storage is proportional to the number of data items and independent of the number of processes that access the data structure.

References (16)

  • W.J. Dally et al.

    The balanced cube: A concurrent data structure

  • C. Ellis

    Concurrent search and insertion in AVL trees

    IEEE Trans. Comput.

    (Sept. 1980)
  • C. Ellis

    Concurrent search and insertion in 2–3 trees

    Acta Inform.

    (1980)
  • C. Ellis

    Extendible hashing for concurrent operations and distributed data

  • C. Ellis

    Concurrency and linear hashing

  • R. Fagin et al.

    Extendible hashing—A fast access method for dynamic files

    ACM Trans. Database Systems

    (Sept. 1979)
  • M. Imase et al.

    Design to minimize diameter on building-block network

    IEEE Trans. Comput.

    (June 1981)
  • H.T. Kung et al.

    Concurrent manipulation of binary search trees

    ACM Trans. Database Systems

    (Sept. 1980)
There are more references available in the full text version of this article.

Cited by (0)

The author thanks Udi Manber for the original problem and many helpful discussions.

View full text