MOSS Forum

 
Home » Forum » MOSS       Ask a questionRSS Feeds

Seeking MOSS2007 Search Advice

  Asked By: Roxanne Horton         Date: Apr 25, 2007      Category: MOSS      Views: 215
 

Here is the situation: We want to index the resumes for searching. For each
resume, the physical resume is located in a file share, and there is a
corresponding row in a sqlserver 2005 database table with other information
about the resume, such as the job seeker'first name, last name, the resume
location (pointing to the file share), email address, current company name etc.
For searching, we would like to have full text search and fielded search too,
such as by last name, current company name etc.

We know that MOSS can index a file share, but how do we combine the info in the
file share and the database together for indexing and searching?

Tagged:        

 

11 Answers Found

 
Answer #1       Answered By: Lynette Sawyer          Answered On: Apr 25, 2007       

You might consider migrating the SQL Server database information into
metadata stored with the files in document properties. This could
probably be automated with a script if you have a large repository. Once
all of the metadata is with the files, MOSS search  will have easy access
to it. Are all of the resumes in the same format? If they're in
different formats, then you would likely need to search / organize on
differing document properties depending upon file  format (which would
not be ideal).

The same approach to migrating the documents to a SharePoint library
(and discontinuing the database) might also be appropriate.

 
Answer #2       Answered By: Shelton Dickson          Answered On: Apr 25, 2007       

The resumes are in different formats, such as MS Word, PDF, etc, so
the option of migrating the mata data to doc properties may not be ideal.

For the second option of migrating the resumes to doc library, can I
setup a search  scope pointing to that library? I tried to do so but
not successful. I created a search scope with the following rule:
Folder = http://crpspt03d/Unsolicited%20Resumes, (I also tried the
follwoing rule: Folder =
http://crpspt03d/Unsolicited%20Resumes/Forms/AllItems.aspx). I
uploaded 5 resumes to the library and the search scope shows 0 items
in the scope. Did I miss something here with regard to setting up the
search scope?

Another option is to develop custom protocol handler, but the
documentation on this is very thin...

 
Answer #3       Answered By: Caleb Gordon          Answered On: Apr 25, 2007       

Documents in a Document Library can be associated with common metadata
eliminating the tangential SQL database as Bradley suggested. SharePoint
will do the hard work. All you need do is upload the documents into a
library that includes the extended metadata attributes you want and then
populate those attributes with the information currently housed in the
SQL database. Shouldn't be a long or difficult process.

 
Answer #4       Answered By: Angarika Shroff          Answered On: Apr 25, 2007       

I like the document library idea and I have created some site columns
for metadata and created a content type named resume, and associated
the resume content type with my document library. Also my new scope is
working now since I need to crawl the default content souce. I was
able to add a new tab to the search  page too. Now I am trying to
figure out the how to modify the property picker dropdown on the
advanced search box webpart.

A slight change of topic, how much data (in terms of number of
items/physical size) can a document library hold? Should I use a
document center intead? For the situation, we have about 700,000
resumes already and growing. What is the impact on performance and
maintenance?

 
Answer #5       Answered By: Micheal Knight          Answered On: Apr 25, 2007       

Here is a link that describes the limits of the old version of the product:
markharrison.co.uk/.../...point-scaling-limits.htm
which is pretty much what the product can do now. The gist is that a
document library can have millions of documents in it, but you will clearly
want to use lots of folders so that each folder doesn't have more than a few
thousand documents in it.

 
Answer #6       Answered By: Lesley Tate          Answered On: Apr 25, 2007       

The capacity planning guidelines for WSSv3 and MOSS2007 are here:

WSS:
technet2.microsoft.com/.../0e13e879-0156-4e9f-b6
d5-abddcf326d211033.mspx

MOSS:
technet2.microsoft.com/.../eb2493e8-e498-462a-ab5d-1
b779529dc471033.mspx

 
Answer #7       Answered By: Chris Daniel          Answered On: Apr 25, 2007       

I would strongly suggest you not place millions of documents in a single
library - not because it cannot be done, but because usability will
suffer greatly.

 
Answer #8       Answered By: Lynn Mann          Answered On: Apr 25, 2007       

Definitely we need to have decent usability. There is only way to find
it out ...

Not sure anybody has any info on custom protocol handler
implementation for MOSS2007?

 
Answer #9       Answered By: Damini Dande          Answered On: Apr 25, 2007       

I don't think that this changed from v2 to v3.

 
Answer #10       Answered By: Addison Peck          Answered On: Apr 25, 2007       

Is the library part of an existing content source and crawl schedule? If
not, it needs to be included.

How often is the content crawled? Perhaps an incremental crawl hasn't
been conducted yet.

 
Answer #11       Answered By: Lalit Bhattacharya          Answered On: Apr 25, 2007       

Let me know how you make out with that
property picker drop-down. We've contracted out a custom web part for
that.

 
Didn't find what you were looking for? Find more on Seeking MOSS2007 Search Advice Or get search suggestion and latest updates.


Your Answer
  • Answer should be atleast 30 Characters.
  • Please put code inside [Code] your code [/Code].