Scott was kind enough to include a reference to a comment I made a few days ago in his post on the subject of caching. Clearly what he discusses goes way beyond my hmmmm..... shall we say "pedestrian" notions of caching. Although if anyone can build the systems he describes, that person ought to be doing rather well in spite of the current downturn in the tech sector.
Speaking of caching....
I am interested in caching web "reports" in some way. A web "report" is simply an aspx page in some sort of easy to print format. By caching, I'm wondering if there is a way to record the criteria a user enters to generate the report as well as store the output generated based on that criteria. Perhaps in an XML file (or some other format de riguer). Conceivably, if a user enters the same criteria again, I could just retrieve his previous results from a file - AS LONG AS THE DB HASN'T CHANGED.
And therein lies my boogaboo. Its a nice concept but what to do if the underlying tables in the db have received data changes. Is there a way to create some sort of notification measure that would signal the processing logic to go back to the db instead of the saved XML file. Perhaps even something along the lines of Scott's discussion, although that almost sounds like overkill for my modest needs. I haven't quite thought this all the way through because it appears some kind of db access is needed no matter what, so at best I maybe able to reduce the processing time and resources that it took to build the report in the first place. Comments and ideas are welcome.
7:36:56 PM
|