You'll need to rely on one of the larger ISP/hosting providers for this, it won't be cheap. Exodus (are they still around?), is an example of a well known provider that can accommodate you. Their NOC's (network operation center) are notoriously renowned as being the biggest and the baddest (including guards armed with AR15's back in the "dot com hay days"). You'll also want to do a little research on your hosting provider's providers - they should have redundant uplinks to at least two of the larger pipes (abovenet, UUNet, gblX, etc). They should provide you with evidence of their uptime (MRTG charting) - as you see downtimes for one routing interface, there should be significant jumps for the others to show that they're in fact picking up the slack. This is what I believe the CCIE certifications were meant to teach people - becoming conversant with the various protocols (OSPF, RIP, HSRP, insert TLA here) and how to configure them to make the Beast happy.
As for keeping your application/database transactions synchronized, that could fill a book in it's own right. You're better off dumping the responsibility onto somebody that's done it before. When everything is set up, you'll want to document and test a disaster recovery plan. Adam. > -----Original Message----- > From: Al Musella, DPM [mailto:[EMAIL PROTECTED]] > Sent: Saturday, May 11, 2002 1:05 PM > To: CF-Talk > Subject: Geographical redundancy? > > > I have to put together a proposal for a medical office management > application which will be used in an ASP (application > service provider - > not the MS language:) model.. one of the requirements is that the > application has to be hosted in such a way that a major > disaster (natural > or otherwise) in 1 location can't cause the loss of any data, > and only a > small (maybe an hour) downtime for the application. > After the Sept. 11 tragedy, my websites had connectivity > problems on > and off for a few days. We also had 24 hours of downtime > when a hurricane > knocked down a bunch of telephone poles near my ISP a few years ago. > For this application, that wouldn't have been acceptable. > > I have no idea how to approach it. Any ideas? > > > Al > a1webs.com > > > > > At 06:08 PM 5/10/2002 -0400, Justin Greene wrote: > >I Have to agree. Hardware based clustering for the front > end... and either > >SQL Enterprise or Veritas on the backend to handle the > database cluster. > >Very solid configuration. We have been hardware clustering > CF with Alteons > >for over 3 years. Just need to keep sessions in the DB and > make sure the > >web boxes keep the file systems synched. > > ______________________________________________________________________ Your ad could be here. Monies from ads go to support these lists and provide more resources for the community. http://www.fusionauthority.com/ads.cfm FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Archives: http://www.mail-archive.com/cf-talk@houseoffusion.com/ Unsubscribe: http://www.houseoffusion.com/index.cfm?sidebar=lists
