Forum Moderators: open
I need to store a tree structure in a DB. It will be far more read from than written to, but still needs to be written to.
It will get quite large (say, 50,000 sub or sub-sub nodes).
I did some research and came across two methods in common use - adjacency list and modified pre-order traversal. The latter, modified pre-order traversal, seems certainly more efficient from a read perspective, but on a large tree updating the right values of all these sub and sub-sub nodes could be time-consuming and slow.
Are there any other methods that I've missed that are suitable for this kind of data hierarchy ?
Many thanks....
That said, 50,000 seems an awful large number for categories and subcategories.
Yes. Indicates perhaps we need to re-think the overal information architecture. Trouble is we keep coming back to tree structures as it's the natural "fit" for the type of data being stored.
I'll try and re-think it again - thanks for the input.
[edited by: coopster at 12:36 pm (utc) on Mar. 19, 2009]
for the type of data being stored
Wait a minute, are we talking data or categorization of data?
