Q: HADR and LTAPEDEV = NUL
Greetings, We have two identical Netfinity servers running NT4/SP3 and Informix 7.30.TC7. Would it be possible for us to use High Availability Data Replication to have one Netfinity act as PRIMARY (accepting updates), the other act as SECONDARY (read-only) and backup logical logs to NUL?? Our server(s) drive an automated shipping conveyor, and it is very important for us to be able to handle any hardware failure quickly, getting the distribution centre back in motion (thus the duplicate machines). Data is very transient, however, so I'm not quite sure about the backup strategy: Being able to restore to 30 minutes before a failure is no more useful than being able to restore to a week before. This line of thinking brought me to the "HADR but backup to NUL" idea. Given our circumstances, would this be a sane and/or reasonable strategy? All help most appreciated, Ken Sent via Deja.com http://www.deja.com/ Before you buy.
In article <7tandj$a5r$1@nnrp1.deja.com>, ksizer@danriver.com writes >Greetings, >We have two identical Netfinity servers running NT4/SP3 and >Informix 7.30.TC7. Would it be possible for us to use High >Availability Data Replication to have one Netfinity act as >PRIMARY (accepting updates), the other act as SECONDARY (read-only) >and backup logical logs to NUL?? > >Our server(s) drive an automated shipping conveyor, and it is very >important for us to be able to handle any hardware failure quickly, >getting the distribution centre back in motion (thus the duplicate >machines). Data is very transient, however, so I'm not quite sure >about the backup strategy: Being able to restore to 30 minutes >before a failure is no more useful than being able to restore to a >week before. > >This line of thinking brought me to the "HADR but backup to NUL" >idea. Given our circumstances, would this be a sane and/or >reasonable strategy? > That is what HADR is about. But take archives+ backup logical logs in case both machines fail. At least you still have the data then! > >All help most appreciated, >Ken > > >Sent via Deja.com http://www.deja.com/ >Before you buy. -- David Williams
In article <$YthpLAyYQ+3EwTq@smooth1.demon.co.uk>, > David Williams <djw@smooth1.demon.co.uk> wrote: > >This line of thinking brought me to the "HADR but backup to NUL" > >idea. Given our circumstances, would this be a sane and/or > >reasonable strategy? > > > > That is what HADR is about. But take archives+ backup logical logs > in case both machines fail. At least you still have the data then! If I have to, I have to, but the "data churn" is such that we'll run through logical logs very quickly and the chance of both machines failing (and the distribution centre not being blown up) is very slim. Am I correct in assuming that I need to either keep all logical logs (in sequence) since the last backup, or none of them? If HADR cannot function with logical logs going to NUL, could I send them to flat files and just not keep them? Thanks, Ken Sent via Deja.com http://www.deja.com/ Before you buy.