server as content source

Question

Does anyone know how to add a content source of a complete server.

ie \\server instead of \\server\share.

I have several servers with a bunch of shares.  Most are people's home folders.

I was hoping I could crawl the whole server, and using permissions only the people with the right permissions would be able to see their files.

 

Thanks Charles

 

 

 

Answer 1

Hi Charels,

 

By default, the file share content source type should folow and crawl sub-folders.  If \\server is the file system parent of \\server\share, then that should be your start address.  If it's something else, e.g. D:\data\users\share on "server", then try making the folder \\server\users a share and using that as the start address.

 

Regarding permissions, as long as the browse permissions are set correctly on the resources, Search Server will crawl them and present security-trimmed results to whomever is searching.

 

Sage

 

Answer 2

 

OK, here is the error that i get from the crawl log. 

 

The item could not be crawled because the Microsoft Single Sign-on service failed to connect to the backend database

server.

 

This error came from the crawl source as \\server. 

 

I just wanted to make sure that you can't put a computer name as a crawl source.  It can only be a share name ie \\server\share. 

I suppose I can try an administrative share, ie.  \\server\d$.

 

Basicly i wanted to crawl an entire server without having to modify or add crawl content on a regular basis.

 

 

 

Answer 3

Hi Charels,

 

I spoke with one of the crawling experts here, and he confirmed the behavior you are seeing is by design.  The crawl error is actually an authentication failure, but even if it did authenticate your crrawling account, it would hit the real issue.) 

 

Basically, the crawler is not able to enumerate shares on a server.  To get around this here, one of the techniques we use here is to arrange all of the shares beneath a common, shared parent.  For example,

 

\\server\shares\ is a share and the start address for the content source.

\\server\share\user1,

\\server\share\user2,

\\server\share\user3 are crawled automatically.

 

The other alternative is manage a content source with a lot of start addresses, which I know you're trying to avoid.  In http crawling, one trick for a similar problem (lots of random, unlinked sites) is to create a single html file with links to the top page of all of the sites, and then crawl that one html file.  You could experiment with a similar workaround, perhaps using a script to enumerate shares and write the file. 

 

Sage

 

 

Answer 4

OK, Thanks.  I just wanted to make sure.