Can your links point to static pages (would mean some redesign work, I'm sure)? If so, can you then physically build static html pages during your once-nightly update and serve up plain .htm -- or static .cfm, if you have to -- files during the day? The former would bypass CF server processing as well as the db.
--Matt Robertson-- MSB Designs, Inc. http://mysecretbase.com -----Original Message----- From: Andrew Peterson [mailto:[EMAIL PROTECTED]] Sent: Thursday, January 17, 2002 9:52 AM To: CF-Talk Subject: RE: Caching Dynamic Pages - better way? (Long) > Just a thought, but since you know that the DB2 database is going to > be updated every night, why not schedule a process that runs > nightly right after the data import Yes - that is a good idea and I think that's where we may be headed if there is no way to eliminate the "active database" altogether. Thanks! Andrew Peterson > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] > Sent: Thursday, January 17, 2002 11:49 AM > To: CF-Talk > Subject: RE: Caching Dynamic Pages - better way? (Long) > > > Just a thought, but since you know that the DB2 database is going to > be updated every night, why not schedule a process that runs > nightly right > after the data import that does a bulk data transfer to your "active > database" (access or whatever)? Then run your queries based on that. > > Hatton > > > ... The data in question is updated once nightly. Since this caching > > mechanism was built long ago, I was wondering if today > there was a better > > way to do what were doing, which is the following: > > > > In order to initially cache the web page, two variables are intially > > captured and inserted into an MS Access database (I know, I > know). Those > > variables are Query and Params. The Query variable contains > the page path > > and name, i.e. /Expert/Revenue/qryGetTotalRev.cfm, and the > Params variable > > contains the name/value pairs, i.e. > > Control=#Control#&Reset=Y&GroupBy=#GroupBy#&SortName=Yes. > > > > Once the new record is inserted into this database, > CFX_HTTPGET is used to > > create the page, display it for the user (just as a browser would), > > and also file it as a html page into a file folder with a unique > > name. > > > > The next time this page is hit, a query is run on the MS Access > > database to see if the page has been cached, and if so, looks up > > which html page contains the data so that next user that visits will > > see > the cached html > > page, rather than hitting db2 to get the data. That cached html page > > is then served to the user using a cfinclude. > > > > I see this solution as having at least two problems. First, > we're still > > hitting a database (an Access one at that), and second, it > adds another > > layer of complexity - maintaining that Access database. I've since > > tried to cache all the queries into memory using CFIDE, but when I > > limit the maximum > > number of cached queries to 500, the cache only lasts about > 20 minutes. > > We've got a Gig of memory on that box, so I'm thinking of > increasing this > > number ten fold to see what happens. Thoughts? > > > > After doing some research, increasing the cached queries > seems to be the > > only viable alternative. Any other ideas on how to approach > this would be > > greatly appreciated. I've not yet done a direct comparison of the > > processing time of the db2 queried pages vs the Access pages, but > > I'm > preparing to do > > just that. > > > > Thanks! > > Andrew Peterson > > > ______________________________________________________________________ Get Your Own Dedicated Windows 2000 Server PIII 800 / 256 MB RAM / 40 GB HD / 20 GB MO/XFER Instant Activation � $99/Month � Free Setup http://www.pennyhost.com/redirect.cfm?adcode=coldfusionb FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/[email protected]/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
