Re: range remove in BTree, or other table storage

From: Sampo Syreeni <decoy_at_iki.fi>
Date: 23 Feb 2007 00:18:21 -0800
Message-ID: <1172218701.316040.119520_at_t69g2000cwt.googlegroups.com>


On Feb 21, 9:15 pm, laurento..._at_gmail.com wrote:

> Is anybody aware of literature about algorithms to remove a range of
> records from a B-Tree, and doing bulk insertion, or better of a table
> storage implementation which provides that sort of things?

Try at least Carey, DeWitt, Richardson and Shekita: Object and File Management in the EXODUS Extensible
Database System; and Stonebraker, Stettner, Lynn, Kalash and Guttman: Document Processing in a Relational Database System. The interesting part is where they use a variant B-tree to implement efficient insertions into the middle of the key space of a sequentially indexed flatfile -- that sort of thing calls for efficient bulk inserts as well.

As for buffering, the neatest concept I've come across to date is probably Arge: The Buffer Tree: A New Technique for Optimal I/O Algorithms. Received on Fri Feb 23 2007 - 09:18:21 CET

Original text of this message