Considerations For a Portal Site with Search Engine Presence
A client of mine is developing a large portal site using Oracle 9i AS and has asked me whether this program will have any effect on the search engine presence that their site already maintains. Can you tell me whether there are any issues I have to consider in optimising this site for the search engines (above the normal considerations) and what limitations the software might bring in this area.
Many thanks indeed for any help you can give.
- Would the database supporting this portal be co-located on the same server as your search engines?
- Would the portal application maybe leverage the same underlying database structures?
- Would the portal application be linked to the search engine application database using linked servers?
If any of these basic scenarios existed, then the questions I would pose would be as follows:
- What is the performance on the existing search engine today? Is it acceptable or not?
- What sort of contention do you see today on your search engine?
- What sort of contention do you see when you run on the portal engine?
- Are the contention resources related or are they ones that could be eliminated by putting objects onto different disks, adding CPUs, memory, etc.?
- What sort of data is being pulled by the search engine vs the portal?
- How are you trying to get at the data via views, direct joins etc.?
Maybe in order to reduce contention and maximize performance it will be neccessary to create some derived tables which are updated on a specified frequency to eliminate unneccessary contention. How update intensive is the portal engine , if there is a lot, you may have concurrent usage issues there.
Not sure if that helps, but good luck.
Lee Gould , President of Rocketrainer (http://www.rocketrainer.com) a revolutionary web based athletic coaching ASP. Lee does database architecture and development consulting for Oracle, Sybase and Microsoft SQL Server. Lee is a co-author of Transact SQL Programming published by O’Reilly and has been a database professional for over 14 years, working on Wall Street for 7 years. Lee holds an MBA from Seton Hall University.