Welcome to BIT320: Databases and Information | MySmartChannels
I was thinking about what Professor Gibson was drawing on the board today in class, and something occurred to me about the reliability of the LINUX4-LINUX2 relationship. Isn't it true that having users go through LINUX4 before they get to the ORACLE software on LINUX2 reduces the reliability of the database? IE if LINUX4 fails, there is no way to access the ORACLE server LINUX2. This would cause the normal user to believe that there is a problem with the ORACLE server, when in fact the only problem is that the messenger server has gone down... It seems like this situation relaxes the computational load on the ORACLE machine at the price of introducing some extra variability into the process. Am I interpreting this situation correctly?
I think that's the point. If linux4 goes down, the data is still available on linux2, so hypothetically you could point linux3 or linux5 at linux2 to get your data. Or if linux3 isn't in heavy use, you could rename it linux4 or point requests to linux4 to linux3, which would then serve as the liaison to linux4. I guess I'm doing a bad job of explaining redundancy, but I have to imagine the concept is supposed to be at work somehow. Does anyone have any thoughts?
Posted by: Tom Campion | October 12, 2004 at 04:41 PM