index my site using index server

[yaniv]

i"m trying to upgrade my site index and i found here:

http://support.microsoft.com/default.aspx?scid=kb;en-us;820105

the article show how to use microsoft indexing system. but the exmple run on txt files, and i want to index my site page (.aspx & .ascx pages that have ids, lets say i have articles.aspx page that end with id- articles.apsx?id=1, ... id=2 and so on), how can index my entire site, with different ids, pages names, forums threads and so on?

can anyone direct me to article on the subject?

 

[MrPaul]

MIS is a file indexing service

As far as I know, the Microsoft Index Server is designed for indexing files, not web pages - you can set it up to index ASP.Net-related files specially (as described here), but indexing URLs within a web site will most likely require navigating the hyperlinks within the pages of the site, and therefore require web crawling (and indexing) software, not file indexing software.

I may be wrong, since I have rarely used the indexing service, but from what I understand it is not suited to web site indexing.

You state that youre trying to upgrade your site index. What indexing software/approach did you previously use?

 

[yaniv]

Re: MIS is a file indexing service

i used sql server "like" queries, but since my site grow up, i allways get overflows...

 

[MrPaul]

Data types

Personally I would suggest fixing the cause of the overflow errors - they are usually caused by using integer (numeric) datatypes which are too small to represent the numbers they need to.

For example, the Int16 or Short datatype can only hold numbers up to 32,767 so if you are using this type and your search queries are returning more results than 32,767, an overflow will occur. By comparison the Int32 or Integer datatype can represent numbers up to around 2 billion which is sufficient for most applications. Beyond this, the Int64 or Long datatype goes up to around 9 quintillion, though in this case it is probably unnecessary.

I expect many website searches will work with some sort of database query similar to your LIKE queries in SQL Server, so it is by no means a bad approach.

Good luck