关于B-tree的构造，我见过的很多书包括文章都是通过线性insert操作完成，http://en.wikipedia.org/wiki/B-tree里面就提到：it is frequently useful to build a B-tree to represent a large existing collection of data and then update it incrementally using standard B-tree operations. In this case, the most efficient way to construct the initial B-tree is not to insert every element in the initial collection successively, but instead to construct the initial set of leaf nodes directly from the input, then build the internal nodes from these. This approach to B-tree construction is called bulkloading.
但这种方法好像事先必须是有序的。
请问，在真实开发环境里面到底是如何构建B-tree索引的？
