On the other hand I tend to get a database error every 2-3 months on some systems. My technique is to use REPAIR to delete the offending entries in the database blocks and then use merge to copy the lost data back from the shadow system. So far I have not managed to correct the actual error with the help of the 'documentation', it's far easier to just delete & copy over. Not helped by the times REPAIR dies with various errors because it can't handle the database error it is attempting to deal with!!!
Colin "Mark Sires" <[EMAIL PROTECTED]> wrote in message news:[EMAIL PROTECTED] > Paul, > > While I also used to be very good (well, at least highly experienced) with > ^FIX (as well as VIEW when necessary), The only occasion I've had to try > ^REPAIR was an rather unfortunate incident, and the database was way beyond > repair (Cache v2.1.6, and it was not a Cache problem that caused it), so I > can't give you any great instructions on it's use. > > I've had Cache in production since 1999 at large, 24x7 sites, and have never > needed it (knocking on large block of wood on shoulders). Based on this, I > can advise that learning ^REPAIR in depth should be a low priority task. > You'd be much better of learning how to maximize the resiliency of the > database, and how to make recovery easy with shadowing. > > Mark > > > > "Paul Kadow" <[EMAIL PROTECTED]> wrote in message > news:[EMAIL PROTECTED] > > Back in my DSM days I used the ^FIX utility more than I wanted to > > but I became good at it. > > > > I am trying to understand the Cache database and have been trying > > to understand the new version of ^FIX, that is ^REPAIR. > > > > Does anyone have written documentation on the ^REPAIR utility? > > I don't want to wait until I am called upon to use it (I hope I > > never will). Yes, I know I can always call Intersystems but I want > > to be prepared if I can. > > > > > >
