I would like to collect feedback on the issue of dealing with static files, I mean those that are usually generated by a robot for a Web application in order to avoid doing some additional process at the hit level and saving some performance issues.

I have a robot application in all my Web applications. This one is responsible for genereated static HTML pages, usually needed on the main page or available on the site margin. This then allows the Web applications to rapidly display pre-generated content. On the main page, for example, the current table associated with the page is the Site table. So, once a user accesses the main page, I already have a loRow reference to the data row of the site table. So, I was wondering if it would be better to stop generating those static files in HTML to be stored on disk and save this pre-generated HTML code into the site row instead. So, for example, if I would have a static HTML file News0000001.html and Photo000001.html, instead of storing then as is, I was thinking of creating a News and Photo fields in the Site table. Then, once the page is accessed, I could then simply drop the content of those fields in the returned HTML instead of reading the HTML code on disk.

One advantage I see would be the elimination of the retry mechanism, which is something we have to verify, as once the robot generates the file, it may well be requested by a hit. If this is in the SQL Server table, this will all be done by SQL Server automatically.

The Site table is not that huge. We are talking about a few hundreds of records. So, adding a varchar field to hold that data might do. Then, we wouldn't have to deal with files on disk and the Static directory.

Any opinion on the issue about pros and cons between those two approaches?